Search

Intensification of agricultural practices is necessary to support a growing global population. Water movement and storage in soils are very crucial for successful intensification of agriculture. They are important for nutrient delivery to plant and overall crop productivity. Climate change and population growth have been predicted to limit water supply especially in arid regions. Therefore, there is an urgent need to proffer solutions that would help maintain or even increase soil moisture retention. The use of biochar for improving soil hydraulic properties is a current and growing area of research. Biochar is a stable form of charcoal gotten from heating natural organic materials in a high temperature and low oxygen process known as pyrolysis. Biochar physical and chemical properties vary due to the pyrolysis process conditions, the type of feedstock, and the ratios of lignin, cellulose, and hemicellulose in the biomass. These change the structure of the biochar and will invariably affect to what extent it can improve soil water retention. To understand how biochar affects soil water properties we must understand the specific characteristics of biochar that influence these changes. Understanding the mechanisms is important for easy prediction of when and by how much biochar will improve soil water properties. The aim of this thesis is to provide new and extensive insight into 4D structural changes of biochar produced from different feedstock and pyrolysis conditions and how these affects soil moisture characteristics of different soil textures. The first aspect of the thesis was to carry out a meta-analysis of the available literature to quantify the effect of biochar on soil hydraulic properties. To enable matching of biochar to soil constraints and application needs, a thorough understanding of the impact of biochar properties on relevant soil parameters is necessary. This meta- analysis of the available literature for the first time quantitatively assessed the effect of not just biochar application, but different biochar properties on the full sets of key soil hydraulic parameters. The review shows that the key factors influencing biochar performance were particle size, specific surface area and porosity indicating that both soil-biochar inter-particle and biochar intra-particle pores are important factors. Next, the role of biochar particle size and hydrophobicity in controlling soil water movement and retention was assessed. Softwood pellet biochar in five particle size ranges was used for the experiment. These particle sizes were tested on two soil types at four different application rates in the laboratory. The results clearly show that both biochar intraporosity (pores inside biochar particles) and interporosity (pore spaces between biochar and soil particles) are important factors affecting amended soil hydraulic properties. Biochar interpores affect mainly hydraulic conductivity; both interpores and intrapores control soil water retention properties. Our results suggest that for a more effective increase in soil water retention of coarse soils, the use of hydrophilic biochar with high intraporosity is recommended. The third aspect of the thesis was to assess the 4D structural changes and pore network model of standard biomass feedstock during pyrolysis. Biochar properties are highly variable and are dependent on the biomass feedstock and production conditions. Despite these variabilities, there has been no study that uses a single sample to study the development of microstructural properties of biomass and biochar through the full range of pyrolysis temperatures. In this study, synchrotron x-ray micro-tomography (SµCT) was used to visualize the internal structure and characterise pore structure of biochar from several feedstock during pyrolysis (50 - 800 °C). The results show a wide range of variation in the pore structural characteristics of the biochar depending on feedstock and pyrolysis temperature, with observed porosity in the range of 7.41 - 60.56 %. The results from this study are not only important for the use of biochar as soil amendments but also a range of other applications relying on biochar pore characteristics, such as biochar as habitat for microbes, water, and wastewater treatments, as an absorbent for the removal of acid gases, or an additive in construction or engineering materials. In addition, the in-depth insights into changes in biomass structure during heating are valuable for research and applications related to fire safety. Finally, the data set from the SµCT was used to produce statistical models enabling the prediction of the effect of biochar on soil water holding capacity from biochar microstructural properties. The results of this thesis, therefore, clearly establish the mechanisms with which biochar improves soil hydraulic properties. It also demonstrates that by appropriately matching microstructural properties of biochar to those of the target soil, it is possible to achieve considerable improvement in soil properties using relatively low application rates of biochar. The relationship between biochar structural properties and its effect on soil moisture characteristics of specific soil texture, is an attractive pathway towards development of precisely tailored biochars aimed to enhance water use efficiency and provide low-dose, high-efficiency benefits.

Projected precipitation changes due to increasing greenhouse gas concentrations exhibit considerable spatial structure with increases in the mid- to high-latitudes, and decreases in the subtropics. In the Equatorial regions, precipitation is pro- jected to increase over oceans, over Africa and over the Maritime Continent, but to decrease in the Amazon basin. The decrease in precipitation over the Amazon basin and increase over the Maritime Continent are both enhanced in response to plant physiological changes. The aim of this thesis is to identify basic controls on tropical precipitation change over ocean and land, with a particular focus on the zonal asymmetry in the Equatorial region and amplification of precipitation changes due to vegetation. We find that soil-moisture limits on evaporation and a teleconnection between Equatorial Africa and the Amazon basin contribute to the zonal asymmetry in rainfall changes over the two regions. When vegetation changes are included, the amplification of the drying over the Amazon basin is largely due to relatively weak local evapotranspiration, providing further evidence that loss of the Ama- zon rain forest could lead to strong decreases in rainfall. Furthermore, we find that tropical precipitation changes over a warm surface temperature anomaly are related to changes in the atmospheric energy budget, and that monsoon-type rainfall behaviour can be obtained even when the thermal inertias of land and ocean are the same, thus partly contradicting the historical view of monsoons as a large-scale land-sea breeze. Our findings are based on idealised Atmospheric General Circulation Model simulations with continental configurations ranging from flat rectangles to realistic continents with topography. Despite their simplicity, the simulations exhibit pre- cipitation changes similar to complex General Circulation Models, indicating that the mechanisms identified here contribute to tropical rainfall changes on Earth.

Soil moisture (SM) is a key state variable in understanding the hydrologic processes, including runoff, infiltration, drought and crop growth. Thus, acquiring accurate SM information has been a priority in hydrology; however gathering such data remains a challenge for many areas of the world with respect to their spatio-temporal aspects. This challenge has contributed to the popularity of using complementary tools such as satellite retrievals and land surface models. This thesis examines multi-source SM data and further enhances their practical use in SM simulation. First, the exponential filter method is introduced to estimate a root zone SM based on satellite surface SM. Subsequently, a cumulative distribution function matching approach is applied, not only to address the inevitable systematic biases between in-situ and satellite SM but also to determine an ideal temporal combination. The performance of each bias-correction group is validated through a cross-validation procedure. However, a major issue in using satellite SM data for practical applications is their coarse spatial resolution. Therefore, a multivariate stochastic SM estimation approach, based on a Gaussian-mixture nonstationary hidden Markov model, is introduced to spatially disaggregate the satellite SM data for multiple locations. It is revealed that the mean correlation coefficient of the proposed model is significantly greater than that of an ordinary regression model. The second part of this thesis focuses on expanding the applicability of SM data in hydrological applications. I introduce a hybrid modelling framework by incorporating SM state variables obtained from the Tank model and multi-satellite sensors via a machine learning based regression technique. The enhanced performance of the hybrid model over a conventional model (the Tank model) is especially apparent in the simulation of low flows; this indicates that even though the overall contribution to runoff prediction is not significant, satellite SM products appear to help capture distinct features of the rainfall-runoff process. ERA-Interim SM data are then employed to identify the spatial and temporal characteristics of agricultural drought stemming from SM deficit. Furthermore, the copula-based Multivariate Standardized Drought Index is exploited to explicitly determine the interdependence and interaction between precipitation and SM deficiency.

Projected increasing temperatures and reduced summer precipitation in the UK raises questions about the sustainability of aspects of the agriculture industry and food security. This study investigates the potential impact of precipitation changes on sugar beet yield in Eastern England. Observations of precipitation was examined for the period 1971-2000 and the expected changes in precipitation were investigated using seven CMIP5 climate models for the historical phase (1971-2000) and RCP45 and RCP85 future scenarios (2021-2050). Three out of the seven models were found to show good agreement with observations but the MOHC ensemble mean was the closest to the observed means and was used for further precipitation analyses. Statistical analysis of the future precipitation changes were performed using the Met Office Hadley Centre (MOHC) model focused on changes between the historical phase and RCP45 and RCP85. Results showed a consistent and significant reduction in annual May-October precipitation under future scenarios. The study then investigated the impact of reduced future precipitation changes on sugar beet yield by applying controlled watering regimes informed by the CMIP5 projections to sugar beet plants in a greenhouse experiment over two seasons - the use of CMIP5 projections in this way is a first. In the first experiment carried out in 2014, a climatological watering regime (i.e. where the total seasonal rainfall for the different scenarios was applied in equal and regular watering events) was applied to the plants, which meant a 16% reduction in precipitation in the "future" category relative to a "control" category. Analysis of the yields indicated a statistically significant reduction in mean tuber wet mass: mean of 360g for the control and 319g for the future (p-value 0.03). This implies a potential yield reduction of 11% by 2050. In the second experiment carried out in 2015, a "realistic distribution" watering regime (i.e. where the total seasonal rainfall is applied in a series of watering events that reflect the analysed sizes and distribution of rainfall events in the different categories), this meant a reduction in precipitation in the months of June (-15.6%), July (-7.7%) and August (-3.7%). This resulted in statistically significant reduction in mean tuber wet mass between control (153g) and RCP85 (113g) with a p-value of 0.01. This implies a reduction of 26% in future yields under RCP85 by 2050. Results in this thesis further show how changes and variation in precipitation are intertwined with changes in soil moisture and yield of sugar beet plants. The findings will enable UK sugar beet farmers to identify potential areas of challenges in order to adapt their management practices to ensure maximum crop yield in future growing seasons. Moreover, from a global perspective, the findings here are also broadly applicable to a variety of agricultural crops in different parts of the world, where changes in yield may have important consequences to food security and food prices.

A major uncertainty in soil carbon studies is how inputs of fresh plant-derived carbon affect the turnover of existing soil organic matter (SOM) by so-called priming effects. Priming may occur directly as a result of nutrient mining by existing microbial communities, or indirectly via microbial population adjustments. Soil type and conditions may also influence the intensity and direction of priming effects. However the mechanisms are poorly understood. The objectives of this study were (1) to investigate how additions of labile C4 substrate affected SOM turnover in two contrasting unplanted C3 soils (clayey fertile from Temple Balsall, Warwickshire (TB) and sandy acid from Shuttleworth, Bedfordshire (SH) using13 C isotope shifts; (2) to investigate the influence of rhizodeposition from plant roots on SOM turnover in the same two soils planted with a C4 grass; (3) to assess an automated field system for measuring soil temperature, moisture and photosynthesis sensitivities of SOM turnover in the same two soils over diurnal to seasonal time scales. I used a combination of laboratory incubation, glasshouse and field experiments. In the soil incubation experiment, I made daily applications of either a maize root extract or sucrose to soil microcosms at rates simulating grassland rhizodeposition, and followed soil respiration (Rs) and its δ13 C over 19 days. I inferred the extent of priming from the δ13 C of Rs and the δ13 C of substrate and soil end-members. There were positive priming effects in both soils in response to the two substrates. In the SH soil there were no differences in priming effects between the substrates. However in the TB soil, sucrose produced greater priming effects than maize root extract, and priming effects with sucrose increased over time whereas with maize root extract declined after the first week. I explain these effects in terms of the greater fertility of the TB soil and resulting greater microbial nitrogen mineralization induced by priming. Because the maize root extract contained some nitrogen, over time microbial nitrogen requirements were satisfied without priming whereas with sucrose the nitrogen demand increased over time. In the glasshouse experiment, I planted C4 Kikuyu grass (Pennisetum clandestinum) in pots with the same two soils. The extent of rhizodeposition by the plants was altered by intermittently clipping the grass in half the pots (there were also unplanted controls) and priming effects were inferred from the δ13 C of Rs and the δ13 C of plant and soil end-members. Unclipped plants in both soils generated positive priming effects, while clipping reduced priming in TB soil and produced negligible PEs in SH soil. Microbial nutrient mining of SOM again explained the observed PEs in this experiment. Photosynthesis was a major driver of priming effects in the planted systems. In the third experiment, I found that the tested automated chamber system provided reliable measurements of Rs and net ecosystem exchange (NEE), and it was possible to draw relations for the dependency of Rs and NEE on key environmental drivers. Collectively, the results contribute to a better understanding of the mechanisms of priming effects and highlight possibilities for further research. The methods developed here will allow high temporal and spatial resolution measurements of Rs and NEE under field conditions, using stable isotope methods to separate fluxes into plant- and soil-derived components. Keywords: Soil respiration, soil moisture, soil temperature, Isotope ratio, maize root, flux chamber, climate change, organic matter, rhizodeposition.

Synthetic Aperture Radar (SAR) has been identified as a good candidate to provide high-resolution soil moisture information over extended areas. SAR data could be used as observations within a global Data Assimilation (DA) approach to benefit applications such as hydrology and agriculture. Prior to developing an operational DA system, one must tackle the following challenges of soil moisture estimation with SAR: (1) the dependency of the measured radar signal on both soil moisture and soil surface roughness which leads to an ill-conditioned inverse problem, and (2) the difficulty in characterizing spatially/temporally surface roughness of natural soils and its scattering contribution. The objectives of this project are (1) to develop a roughness measurement method to improve the spatial/temporal characterization of soil surface roughness, and (2) to investigate to what extent the inverse problem can be solved by combining multipolarization, multi-incidence, and/or multi-frequency radar measurements. The first objective is achieved with a measurement method based on Structure from Motion (SfM). It is tailored to monitor natural surface roughness changes which have often been assumed negligible although without evidence. The measurement method is flexible, a.ordable, straightforward and generates Digital Elevation Models (DEMs) for a SAR-pixel-size plot with mm accuracy. A new processing method based on band-filtering of the DEM and its 2D Power Spectral Density (PSD) is proposed to compute the classical roughness parameters. Time series of DEMs show that non-negligible changes in surface roughness can happen within two months at scales relevant for microwave scattering. The second objective is achieved using maximum likelihood fitting of the Oh backscattering model to (1) full-polarimetric Radarsat-2 data and (2) simulated multi-polarization / multi-incidence / multi-frequency radar data. Model fitting with the Radarsat-2 images leads to poor soil moisture retrieval which is related to inaccuracy of the Oh model. Model fitting with the simulated data quantifies the amount of multilooking for di.erent combinations of measurements needed to mitigate the critical e.ect of speckle on soil moisture uncertainty. Results also suggest that dual-polarization measurements at L- and C-bands are a promising combination to achieve the observation requirements of soil moisture. In conclusion, the SfM method along with the recommended processing techniques are good candidates to improve the characterization of surface roughness. A combination of multi-polarization and multi-frequency radar measurements appears to be a robust basis for a future Data Assimilation system for global soil moisture monitoring.

Artificial Intelligence (AI) has advanced rapidly in the past two decades. Internet of Things (IoT) technology has advanced rapidly during the last decade. Merging these two technologies has immense potential in several industries, including agriculture. We have identified several research gaps in utilizing IoT technology in agriculture. One problem was the digital divide between rural, unconnected, or limited connected areas and urban areas for utilizing images for decision-making, which has advanced with the growth of AI. Another area for improvement was the farmers' demotivation to use in-situ soil moisture sensors for irrigation decision-making due to inherited installation difficulties. As Nebraska has a unique agricultural industry, including 8.6 million irrigated croplands accounting for 14.8% of all irrigated croplands in the United States and a significant amount of under-connected farmlands, we decided to address these two pressing issues. Chapters three and five discuss the design and development of the AICropCAM platform. AICropCAM can extract useful information such as crop type, canopy coverage, and pest pressure and transmit it to cloud servers through low-power, long-range communication mediums using deep learning models. This platform can reduce the digital divide in crop monitoring digital solutions. Chapter four discusses an effort to predict the soil water properties, including soil water content, soil water deficit, and remaining water content, through machine learning models that depend on the above canopy sensors and real-time weather data. The current models were trained using 2022 and 2023 data, showing promising prediction accuracy above 95%. However, these models must be tested in the long run for commercial use. At the top level, the case studies discussed in this dissertation will help reduce the rural-urban gap and digital divide in agriculture. Additionally, the proposed AICropCAM and the above-ground sensor use for soil water property estimation can increase pivot irrigation systems' crop monitoring and improve irrigation decision-making tools. Ultimately, the case studies in this dissertation conclude that the integration of AI and IoT can advance the crop monitoring practices used for precision agriculture in mid-western irrigated cropping systems. Advisor: Yufeng Ge

Water colour is the naturally occurring yellow-brown 'tea like' discolouration which can be observed in freshwaters, and is typically composed of high molecular weight organic carbon. Water discolouration is a major problem for the water industry as over the past 30 years, water colour release has more than doubled, greatly increasing treatment costs and making land management a more feasible option for improving water quality. This project was developed in conjunction with Yorkshire Water Plc, to investigate the mechanisms of water colour production, and the implications that these mechanisms have for land management to improve water quality. In particular, the study aimed to identify the key relationships between drainage water quality and quantity, soil processes and vegetation type that are pivotal to the understanding of water discolouration.

This dissertation is a three-part interconnected geographic study of fine scale soil moisture using monitoring, modeling, and remote sensing in coastal northern California. Although the focus is on hydrology and soil moisture, soil moisture naturally lends itself to be a unifying theme in physical geography connecting hydrology, agriculture, ecology, climate, wildfire, remote sensing, and humans through management actions. The objective of the three chapters is to advance the characterization and process understanding of soil-plant-water dynamics spatially and temporally and to provide high resolution soil moisture data that can be used by decision makers to improve situational awareness and planning through streamflow forecasting, landslide detection, drought warning and fire danger.In Chapter 1, monitoring data was collected in the context of a paired watershed study and soil and tree moisture probes were used to examine soil water-tree-ecosystem dynamics in Sonoma County, California. The results of this study showed: 1) there were noticeable effects on soil moisture and evapotranspiration from mechanical tree thinning and prescribed burns compared to an untreated watershed, 2) soil moisture decreased at the surface and increased in the root zone post-treatment, relative to an untreated watershed, 3) evapotranspiration decreased by 90 mm in the post treatment year for the treated watershed, 4) soil and tree moisture in situ sensors can be used as a surrogate for discrete and time consuming Live Fuel Moisture (LFM) measurements, and 5) soil moisture thresholds can be used to help identify wildfire danger and drought conditions. Improving real-time wildfire danger and drought conditions using existing or low-cost in situ sensors can improve situational awareness and planning. In Chapter 2, a novel L-band Unoccupied Aerial System (UAS) was used to map surface (0-10 cm depth) soil moisture at a high resolution in a complex, natural landscape. L-band UAS was used to map soil moisture at 5-50 meters resolution in a 3,200-acre mixed grassland-forested landscape in Sonoma County, California. The L-band dielectric constant model produced soil moisture maps with an unbiased root mean squared error (ubRMSE) of 0.07 m3/m3. A random forest machine learning model was also applied and produced surface soil moisture maps at 10-m and 30-m using optical UAS and thermal infrared (TIR) satellite data with accuracy levels of 0.020 m3/m3 and 0.017 m3/m3 ubRMSE, respectively. Improved fine-scale soil moisture maps developed using UAS-based systems may be used to help reduce fire risk and improve hydrologic models, streamflow forecasting, and early detection of landslides. In Chapter 3, high resolution root zone soil moisture, live fuel moisture, and fire danger maps were developed using machine learning, water balance, and regression models that connected the results from Chapter 1 and Chapter 2. Root zone soil moisture maps were developed using a random forest model that resulted in R2 values from 0.71 to 0.80. LFM maps were made using relationships developed in Chapter 1 and modeled soil moisture anomalies with an R2 of 0.82 compared against LFM data. The soil moisture anomaly and LFM results were applied to produce retrospective maps of fire danger that indicated predominantly medium to high fire danger during two historical fire ignition dates. An accumulated soil moisture anomaly (aSMA) showed interannual and long-term patterns and characterized an above-normal aSMA that switched to a below-normal aSMA right before the large wildfires that occurred between 2015-2020. Chapter 1 showed that root zone soil moisture was a better predictor of LFM than surface soil moisture and in Chapter 3, root zone soil moisture anomalies were shown to be an accurate predictor for LFM maps that can be useful to assess wildfire danger conditions at a high temporal and spatial scale. High resolution fire danger and LFM maps can be used to improve awareness of landscape conditions and early drought warning for decision makers.

A main constraint to maize production in New Zealand is yield variability due to the low rainfall supply and also the erratic distribution of rain in the summer which consequently affects crop N uptake and utilization. The aim of this study was to understand the influence of N and water on canopy development, crop growth, phenological development and yield formation. Crops with different yield potentials were created using different levels of N and water availability. Two experiments were carried out, marking the growing season, Experiment 1 in 2015/16 and Experiment 2 in 2016/17. The two experiments were carried out at two different locations at Lincoln University, Canterbury, New Zealand. Experiment 1 was arranged in a split-plot design with four nitrogen (N) levels under different water regimes. The N levels were N1- nil, N2 -75 kg N/ha, N3 – 150 kg N/ha and N4 – 300 kg N/ha, and water levels at Irr1- Irr4 as defined by the accumulated potential soil moisture deficit as 443 (rainfed), 367, 301 and 226 mm, respectively. These created grain yields that were not different in all the treatments which averaged 12.4 t/ha and only varied in total dry matter (DM) accumulated. To create more distinct differences in grain yield, this experiment was repeated at another location with a higher dose of N. Experiment 2 used two levels of water and N in a randomised block design. The treatments were for N1 –nil N and N2 - 500 kg N/ha, rainfed and irrigation (accumulated potential soil moisture deficit at 536 and 296 mm respectively). In Experiment 1, grain yield was not different across the crops and average 12.4 t/ha. The total DM was 19.1 t/ha for the rainfed crop (Irr1) and averaged 22.4 t/ha for the irrigated crops. Grain yield and total DM as explained by intercepted photosynthetic radiation (iPAR) accumulated to 570 MJ/m2 for Irr1 and was higher at 1082 MJ/m2 for Irr2. All crops GAI reached a maximum at 3.7 m2/m2 at a rate of 0.01 m2/m2/°Cd in the duration of 677 °Cd which justified the similarities in grain yield. After the linear phase, the intercepted light in Irr1 immediately reached an asymptote as the leaves withered quickly due to water stress. The radiation use efficiency (RUE) was 2.21 g/MJ for N1 and 2.49 g/MJ for N4 and mainly because of the specific leaf N (SLN). The SLN was highest at 2.1 g N/m2 for the irrigated crop with N and lower at 1.66 g N/m2 without N. The contribution of SLN to yield was reflected in total DM. In Experiment 2, grain yield increased progressively from 0.98 t/ha under rainfed to 9.0 t/ha when irrigated and further to 16.3 t/ha with N. Total DM followed a similar response with rainfed accumulating only 4.10 t/ha and 14.3 t/ha under irrigation and doubling with N, creating total DM of 28.9 t/ha. The difference in total DM was explained by the differences in the total amount of iPAR. Under rainfed the total iPAR was 448 MJ/m2 and increased to 551 MJ/m2 with N, and when irrigated was 816 MJ/m2 and further increased to 1005 MJ/m2 with N. The rate and duration changed, indicating the capacity of the crop to capture light depended on the changes in pigment protein complexes, directly linked to development of the GAI as a process of leaf development and expansion. The maximum GAI was affected by the main effects of water and N where GAI increased from 2.14 to 3.49 m2/m2 with water and from 2.48 to 3.14 m2/m2 with N application. In the contribution of RUE to grain yield, SLN was a key factor connecting leaf N concentration to DM production. The SLN was 1.23 g N/m2 at 905 °Cd in all the crops, however, dissecting the canopy into cohorts, SLN varied. The main section of the canopy that supplied assimilates directly to ear development was the mid-cohort. This cohort was affected by both water and N, increasing SLN from 1.43 to 2.39 g N/m2 with water and from 1.46 to 2.37 g N/m2 with N. The changes at cohort levels were explained by GAI as parameter relating to canopy development and the allocation of N within the leaf, and light penetration through the hierarchical canopy arrangement. The amount of water used to produce the given yield in Experiment 2 depended on crop water use which was converted to WUE. Only the crop under irrigation and N was efficiently converting water to DM at 47 kg DM/mm of water. The WUE for the rainfed crops was 17.9 kg DM/mm and did not differ from the irrigated crop without N which had a WUE of 25.7 kg DM/ha/mm. Leaf and canopy photosynthesis has led to improvements in crop DM and yield through enhanced DM partitioning. In depth understanding into elements of SLN is essential for estimation of the SLN throughout the cropping season. The future of crop improvement strategies is dependent on maximising the leaf and canopy photosynthesis and converting DM accumulation into yield benefits.

This thesis examines the hydrological processes of a restored coastal dune slack at St. Fergus, north-east Scotland. The fundamental hydrological processes of the saturated and unsaturated zones of these habitats is relatively poorly understood yet these ecosystems are under increasing pressure from industrial and recreational development and the possible impact of climate change. Within a 1.2km2 catchment, a detailed field investigation was initiated at differing spatial scales to examine; (i) the spatial and temporal variability in soil moisture retention and movement through the unsaturated zone; (ii) the behaviour of the water table and; (iii) the main hydrogeochemical processes over the year October 1996 to September 1997. The field research was augmented with the development of a groundwater flow model of the site which was used in a predictive role to examine the possible implications of future climate change on the groundwater regime of the dune slack and the impact of this on the ecology and geomorphology of the dune system. The field and modelling results were integrated to provide a relatively holistic interpretation of the hydrological processes of the dune slack system. The soils of all study sites were relatively similar in their retention and temporal variability of soil moisture over the year. Soil moisture contents were generally very low, which was predominantly related to the low organic matter content and highly porous sediment. These soil properties resulted in great temporal variability in soil moisture content over the year with no pattern of higher retention in the winter and lower in the summer. Groundwater behaviour was also relatively homogeneous in the catchment, though local scale, transient processes and phenomena appeared to be important. The main influences on groundwater levels and direction of flow included; (i) the groundwater recharge processes (influenced by variable specific yield phenomena, preferential flow paths and the height of the capillary fringe); (ii) the hydrogeology of the site, particularly the complex plateau region; (iii) site topography; (iv) the shallow water table and; (v) terminal encroachment on to the dune slack. The modelling procedure highlighted the need for an integrated field and modelling approach as problems encountered were resolved through interpretation of the field data. Issues raised focused on scale and scaling in groundwater modelling, the validation of such models and the importance of good spatially distributed field data.

Using data collected from six headwater peatland catchments at the Marcell Experimental Forest in northern Minnesota, I assessed the relationship between variability in annual precipitation and annual changes in catchment water storage. Three hypotheses are addressed; (1) annual variability in precipitation is a primary driver of catchment storage change, (2) years of below average precipitation drive the relationship between precipitation and catchment water storage change, and (3) winter and fall precipitation variability are significant seasonal drivers of the annual catchment water storage change. The above relationships were analyzed via cross-correlation lag analysis and linear regression analysis of long-term precipitation, peatland water table elevation (WTE), and upland soil moisture (SM) time series, where WTE and SM served to quantify catchment water storage. Results indicate strong correlations between annual water storage change and annual precipitation variability, both in contemporaneous and antecedent years. Concurrent fall precipitation and antecedent winter precipitation were found to have the most influence on a given year’s water storage change. Years in which precipitation fell below the catchment average (dry years) exhibited a moderately significant linear relationship with annual catchment water storage change. Results of the above analysis were used to create a series of multivariate linear regression models, both with and without moving-average (MA) errors; these models were able to explain between approximately 50% and 70% of the variance found in the annual water storage change time series. Boreal peatlands play a vital role in the planet’s carbon cycle; developing a better understanding of the hydrologic function of these environments will likely prove important to future climate management practices.

Hazard analyses of rainfall-induced landslides have typically been observed to experience a lack of inclusion of measurements of soil moisture within a given soil layer at a site of interest. Soil moisture is a hydromechanical variable capable of both strength gains and reductions within soil systems. However, in situ monitoring of soil moisture at every site of interest is an unfeasible goal. Therefore, spatiotemporal estimates of soil moisture that are representative of in-situ conditions are required for use in subsequent landslide hazard analyses. This study brings together various techniques for the acquisition, modeling, and forecasting of spatiotemporal retrievals of soil moisture across areas of Eastern Kentucky for use in hazard analyses. These techniques include: A novel approach for determination of satellite-based soil moisture retrieval correction factors for use in acquisition of low orbit-based soil moisture retrievals in site-specific analyses, unique spatiotemporal modeling of soil moisture at various depths within the soil layer through assimilation of satellite-based and land surface modeled soil moisture estimates, and the development of a novel workflow to effectively provide 7-day forecasts of soil moisture for use in subsequent forecasting of landslide hazards. Soil moisture retrieved through the previous approaches was implemented within landslide hazard and susceptibility analyses across known rainfall-induced landslides within Eastern Kentucky. Investigated analyses were conducted through a coupling of spatial soil moisture retrievals with that of site-specific geomorphologic data. These analyses proved capable in the detection of incipient failure conditions indicative of landslide occurrence over these known investigated slides. These soil moisture-based analyses show that inclusion of soil moisture, as hydromechanical variable, yields a more capable hazard analysis approach. Additionally, these analyses serve as a means to gain a better understanding of the coupled hydro-mechanical behavior associated with the initiation of rainfall-induced landslides.

Satellite remote sensing techniques provide us an unprecedented opportunity to monitor global land surface water status, the outstanding progress of which has been witnessed in the past three decades. In comparison with traditional optical remote sensing with visible, near-infrared, and/or short-wave infrared sensors, microwave bands could provide an all-sky capability for accurate remote sensing of hydro-meteorological states. To measure global surface soil moisture (SM) and vegetation optical depth (VOD), NASA's Soil Moisture Active Passive (SMAP) satellite mission has been launched on 31 January 2015 using L-band (1-2 GHz) radiometry, which could provide corresponding measurements with a global revisiting time of 2-3 days. The retrieval of SM and VOD at L-band often relies on the inversion of the zeroth-order radiative transfer model that can well formulate the emission interaction between soil and canopy layer. However, due to the limited number of observations at vertical and horizontal polarization channels, inverting the model is typically an ill-posed problem. For improved retrievals of SM and VOD, this thesis strives to advance existing inversion algorithms in terms of the following aspects: (1) narrowing down the feasible ranges of SM and VOD retrievals using the information of soil texture and VOD climatology; (2) accounting for the slow temporal changes of VOD through a Sobolev-norm regularization; (3) incorporating the information of nearby spatial observations considering the theorem of spatial autocorrelation that states neighboring geographical elements tend to be similar on a map; (4) combining the merits of multiple SM products through a deep neural network. The results suggest the proposed strategies could effectively reduce the uncertainties in the retrievals and SM and/or VOD beyond existing SMAP official products.

Dissected, loess-mantled terraces (downlands) are a dominant component of the soil-landscape in Southland, New Zealand and commonly contain soils with dense subsoils that impede drainage. Consequently mole and tile drainage systems, which removes excess soil moisture, are widespread. Alongside agronomic benefits, mole and tile drainage creates problematic secondary effects, including altered stream hydrographs, contaminant transfer into surface water and possible modification of soil properties. Today, water quality and quantity management rely heavily on environmental modelling. Models of soil – water interactions are informed by knowledge of soil hydraulic properties and processes, which is gained from a variety of disciplines, including pedology, soil physics and hydrology. Understanding the properties and processes relevant to mole and tile-drained, slowly permeable loess soils is, thus, criticical for reliable simulations of their water dynamics. Concerningly, this understanding remains elusive. There is a need to better understand the variability in soil properties that regulate water storage, movement and connectivity, as well as the intrinsic and extrinsic controls on water flow pathways. These research gaps were addressed by characterising and monitoring a small, mole and tile-drained basin on a typical Southland sheep farm. Surveying of the soil and loess mantle found vertical differentiation of soil properties was due to contemporary pedogenesis, and soil features inherited from buried loess sheets were important sources of heterogeneity in water-regulating soil properties. Catchment side slopes had greater soil variation relative to the interfluves. Buried loess sheets, which had been exhumed by hillslope procceses, were contributing to the subsoils on the basin slopes whereas the soils on the interfluves were formed solely into the uppermost loess sheet. The findings confirm that understanding the soil-geomorphology of loess-mantled downlands is important for improving soil – landscape models, characterising soil spatial variability and producing more accurate soil maps. Ground penetrating radar (GPR) was effective at identifying and mapping mole channels, thus allowing drainage density to be quantified and the mole network to be characterised. It demonstrated that the mole networks had non-systematic patterns and most likely were multi-generational. Ground truthing during the GPR survey showed mole channels were in good condition even after 30 + years without maintenance. From a survey of soil hydraulic properties it was apparent that the mole fracture network created when moles were installed had not persisted. The finding has implications for hydrological and water quality modelling, as preferential flow via the fracture network is commonly considered to be the major route for rapid water and contaminant transfer into moles. Mole channels were not found to influence soil properties directly; however, it was discovered that surface saturated hydraulic conductivity (Ksat) was controlled by antecedent soil moisture state in parts of the landscape where perched water tables (PWT) persisted. This was hypothesised to occur by priming of the macropore network when capillary rise from a PWT removed air pockets, signalling a potential indirect influence of drainage systems on Ksat. A water balance analysis refuted a common assumption of negligible deep drainage in slowly permeable loess, with or without artificial drainage. Deep drainage varied between events as well as seasonally and occurred via piston flow through the fragipan and underlying loess mantle. Runoff (tile and overland flow) responses to rainfall events could be discriminated by thresholds of precipitation volume and/or intensity, depending on antecedent soil moisture state (dry or wet). Infiltration-excess overland flow was the most common form of surface runoff, consistent with the low surface infiltration rates estimated from infiltration experiments. A corollary is that shifts in the distribution of precipitation intensity with climate change could change the overland flow component of the water balance of this landscape. Tile flow from small events was predominantly sourced from the interfluves and upper hollow. As events became larger, the source area expanded until most of the catchment was contributing to tile flow. When considering the mitigation of contaminant transfers to surface water, winter and spring should be be key time periods, and the upper hollow and interfluves key locations, on which to focus. Soil moisture spatial patterns appeared to be influenced by both PWTs and the artificial drainage network and were not well predicted by the topographic wetness index (TWI). In the dry soil moisture state, upslope areas were hydrologically disconnected. With increasing catchment wetness, hydrological connectivity expanded from the lower hollow to the PWTs at the mid- to upper hollow, and eventually up to the interfluve PWTs. Once the hydrological connection between the lower hollow and interfluve had been established, the catchment had completed its transition into the wet state, and the interfluve PWT and mole network became the primary regulator of tile flow. The transition from dry state to wet state soil moisture conditions took about two weeks and began at the lower catchment hollow. Only rainfall periods with exceptional precipitation volumes and intensities connected the slopes, and only when all other parts of the catchment were saturated. There was no evidence that natural interflow, either as matrix or preferential flow, led to hydrological connection. Surface-sourced preferential flow to the tile drain occurred when the soil was in the dry state, but the flux was hydrologically insignificant. The thesis highlights: The need for significant sources of soil variation associated with loess stratigraphy to be represented in soil classification and mapping; that, while there was no un-drained control in the experimental set-up, the effects of the artificial drainage network appear to be profound and must be considerd in hydrological-contaminant modelling; and finally, that several common assumptions about flow pathways (i.e., mole channel fracture networks, negligible deep drainage, significant natural interflow, insignificant overland flow) in artificially drained, low permeability loess may be inappropriate and, consequently, their use in modelling in support of management and mitigation needs careful evaluation.

Precision agriculture utilizes sensor networks to inform growers on optimal conditions for applying agronomic inputs (water, fertilizer, pesticides, etc.). Precision irrigation is a subset of precision agriculture that focuses on optimizing water usage. While some growers have embraced these techniques, their usage is far from universal due to cost and labor barriers. Several contemporary works have introduced robotic means for crop monitoring and harvesting. These autonomous systems typically use aerial means for sensing to cover broad regions of crops and ground-based rovers for harvesting. Yet, considerably less work has been performed on using ground-based systems to sample and perform direct measurements. Towards this purpose, two robots are designed and tested. The first robot for proximal soil sensing uses an ECa sensor to generate soil moisture maps of an orchard. The second robot for physical sampling uses a 6-DOF robotic arm to pick a leaf from an avocado tree for stem water potential analysis. These two systems present steps toward a larger robotic system for irrigation measurements in orchard crops.

Soil moisture is essential for many applications, such as agricultural irrigation, water resources management, and natural disasters, such as landslides and droughts. With the advancement of measurement technology, a vast amount of soil moisture data is available from ground-based sensors and remote sensing. How can we use such abundant data in a meaningful way? For better interpretability and extrapolation capability, soil moisture data should be analyzed based on a known physical model through an inverse modeling framework. In the dissertation, I explored the inverse modeling of soil moisture dynamics based on the Richardson-Richards equation (RRE) via techniques recently developed in applied mathematics. In Chapter 1, a general introduction is presented. Chapter 2 investigated the application of a neural network-based inverse method called physics-informed neural networks (PINNs). I demonstrated that PINNs with domain decomposition could approximate the solution to the RRE for layered soils by comparing PINNs with an analytical solution of the RRE. Chapter 3 estimated soil hydraulic properties from upward infiltration experiments using the Peters-Durner-Iden (PDI) model. I demonstrated that the PDI model better captured soil moisture dynamics for dry conditions than the van-Genuchten Mualem model. Chapter 4 discusses the estimation of surface water flux from soil moisture measurements through inverse modeling. I compared an adjoint method with PINNs and demonstrated through numerical examples that both methods gave reasonable estimates of surface water flux from soil moisture measurements. However, the adjoint method was more robust than PINNs regarding the reconstructed soil moisture profile for a data-limited case. In Chapter 5, I summarized the limitations of the current approaches and discussed future perspectives of inverse modeling of soil moisture dynamics.

Approximately 30% of the Earth’s land surface is characterized as arid or semiarid, including much of the western United States. Accurate runoff predictions are important for informed watershed management, particularly in rapidly urbanizing areas. Infiltration excess overland flow is the dominant mechanism for runoff generation in many dryland basins, and event-based infiltration or loss models are commonly used to estimate runoff. However, predictions are associated with considerable uncertainty due to the role antecedent soil moisture, an initial condition that must be set by the modeler. The objectives of my research were to (1) evaluate the impact of antecedent soil moisture on runoff, (2) investigate methods for monitoring soil moisture conditions at the catchment scale, and (3) quantify the connection between soil moisture and model parameters to improve model performance. Research was conducted in central New Mexico and based on data from the Walnut Gulch experimental watershed in Arizona. Results under objective 1 show that across spatial scales, soil moisture has a large impact on runoff. High initial moisture content led to substantially higher runoff ratios at the test plot (2.8 m2) and catchment scales (2.8 km2). Failing to account for antecedent conditions caused simulated runoff volume errors of up to one order of magnitude. Given the importance of antecedent conditions, I developed a novel method for monitoring soil moisture conditions at the catchment scale under objective 2. The approach is based on the antecedent precipitation index (API) and uses rainfall estimates from ground-based weather radar. Compared to satellite-derived soil moisture data, the API-based method performed well with respect to accuracy (root mean square error 0.014-0.018 m3 m-3), latency, as well as spatial and temporal resolution. Under objective 3, I optimized parameters for seven simple loss models at the plot scale (2.8 m2) and established relationships between model parameters and antecedent conditions. I then tested model performance at the hillslope (1.5-3.7 ha) and catchment scales (2.4-2.8 km2) based on measured runoff data. At the hillslope scale, rainfall simulation can be used successfully to parameterize models, a valuable tool in data-spares regions. At the catchment scale, most models showed positive bias, indicating that other processes (such as channel- or transmission losses) play an important role in determining the catchment runoff response.

Depending on one’s perspective, surface properties and processes can be either a source of noise or the target of remote sensing observations of the Earth. Interferometric Synthetic Aperture Radar (InSAR), a satellite-borne geodetic method that is used to constrain deformation of the Earth’s surface, is one such remote sensing technique that is impacted by dynamic surface states. In this dissertation, I view surface processes from both the perspective of a noise source to be mitigated in studies of ground deformation, and an independent measure of surface dynamics. In particular, I quantify the impact of mid-time series forest disturbance on InSAR estimates of secular and time variable ground displacement, particularly in regions that experience a significant degree of clearcutting, such as the American Pacific Northwest and Sumatra. I show how the contribution of forest disturbance can introduce either random error or systematic bias into time series estimation depending on satellite geometry, and propose a method for mitigating the contribution of forest disturbance using a priori knowledge of tree loss with independent, optically derived datasets. I then turn my focus to the impact of soil moisture variability on InSAR phase. First, I present a method to invert a dataset derived from a measure of InSAR data quality, known as coherence, that separates different causes of data quality variation over time and in hyper-arid regions where little to no vegetation is present. One product of this inversion is a relative soil moisture metric at each data point in time, and I compare this metric with a variety of other remote sensing and data assimilation soil moisture products over the southern Arabian Peninsula following two extreme rain events in 2018. I find that the coherence-derived metric agrees well with other datasets, but the timescale for which the moisture signal is present after a rain event is significantly longer than other datasets, suggesting that the coherence metric is sensitive to a different part of the soil moisture signal. Finally, I shift focus and analyze the impact of soil moisture on InSAR phase, with the goal of modeling and mitigating the increased amount of phase scatter caused by soil moisture variability. This dissertation presents important advances towards the collective aim to fully characterize how land cover and ground characteristics contribute to InSAR phase.

The promoters of the large groundwater developments implemented in the 1970's paid little attention to the effects of pumping on soil moisture. A field study, conducted in 1979 in the Tern Area of the Shropshire Groundwater Scheme, revealed that significant quantities of the available moisture could be removed from the root zone of vegetation when drawdown of shallow watertables occurred. Arguments to this effect, supported by the field study evidence, were successfully presented at the Shropshire Groundwater Scheme public inquiry. The aim of this study has been to expand the work which was undertaken in connection with the Shropshire Groundwater Scheme, and to develop a method whereby the effects of groundwater pumping on vegetation can be assessed, and hence the impacts minimised. Two concepts, the critical height and the soil sensitivity depth, formulated during the initial work are at the core of the Environmental Impact Assessment method whose development is described. A programme of laboratory experiments on soil columns is described, as is the derivation of relationships for determining critical heights and field capacity moisture profiles. These relationships are subsequently employed in evaluating the effects of groundwater drawdown. In employing the environmental assessment technique, digitised maps of relevant features of the Tern Area are combined to produce composite maps delineating the extent of the areas which are potentially sensitive to groundwater drawdown. A series of crop yield/moisture loss functions are then employed to estimate the impact of simulated pumping events on the agricultural community of the Tern Area. Finally, guidelines, based on experience gained through evaluation of the Tern Area case study, are presented for use in the design of soil moisture monitoring systems and in the siting of boreholes. In addition recommendations are made for development of the EIA technique, and further research needs are identified.

Abstract: Agricultural activities contribute greatly to greenhouse gas (GHG) emissions, accounting for 14% of the total anthropogenic emissions of GHG such as nitrous oxide (N2O), carbon dioxide (CO2), and methane (CH4). Of these GHG, N2O is the most concerning gas because of its high global warming potential (GWP), 298 times higher than that of CO2, and its capacity for depleting stratospheric ozone. The increase of nitrogen fertilizer additions over the years has made agricultural soils responsible for around 60% of global anthropogenic N2O emissions. Biosolids are by-products from sewage treatment processes that can be land applied to agricultural soils in order to recycle their nutrients, improve soil properties and decrease the dependence on commercial fertilizers; however, GHG can be released from this practice. Thus, there is a need to understand the effects of biosolid additions on N dynamics, nutrient use efficiency, crop productivity, and the amount of GHG emissions released by this practice. In a field study, we evaluated the fluxes of N2O, CH4 and CO2, soil available N, barley (Hordeum vulgare L.) biomass productivity, and nitrogen use efficiency (NUE) in croplands receiving three types of biosolids (mesophilic anaerobic digested [BM], composted [BC], and alkaline-stabilized [BA]) and granular urea in a Black Chernozem soil in Central Alberta, Canada, over three experimental site-years. The combinations of each biosolid with urea was also evaluated. All N source treatments were assessed in two placements: surface (S) and incorporation (I) to 15 cm soil depth. Nitrous oxide emissions were triggered by concurrent increases of soil moisture and available N, and incorporation of the N source increased N2O emissions compared to surface-applied N. Annual N2O emission factor (EFarea) from urea-amended soils (0.62 ± 0.14%) were fivefold higher than those from soils receiving only BA or BC (0.12 ± 0.04% or 0.12 ± 0.03%, respectively, P < 0.05), but EFarea from soils amended with only BM (1.33 ± 034%) was more than double the EFarea from urea-amended soils (P > 0.05). Carbon dioxide (CO2) fluxes generally followed similar patterns as the N2O fluxes, while CH4 fluxes were minimal. Overall, the mesophilic anaerobic digested under incorporation treatment (BMI) showed the highest GHG emissions. Results of a partial GHG balance showed that N2O emissions were the main contributor (up to 96%), while urea manufacturing contribution to the GHG balance was up to 52%. This offset the comparatively low field N2O emissions from the urea-amended fields, leading to even higher CO2 equivalents than the BA- and BC-amended fields. Incorporating the N sources enhanced barley biomass, and in certain cases, the combinations of biosolids and urea (e.g., BMURI, BMURS, BCURS) showed even higher biomass and NUE, as well as lower N2O emissions than biosolids-amended soils. Moreover, in an incubation study, we examined the effect of moisture (i.e., 28, 40, 52, and 64% WFPS) and the different types of biosolids aforementioned (i.e., BM, BA, and BC) on N2O production in the referred Black Chernozem soil. We found how the different biosolids properties and soil water contents influenced soil available N dynamics to produce N2O emissions. BM- and BC-amended soils were the higher N2O emitters, and emissions increased with soil moisture. These biosolids-amended soils also showed higher nitrification rates than BA-amended soils and the controls. The NO3−–N concentration by the end of the experiment was well correlated with the total N2O production (r = 0.91). In addition, we examined the sources and priming of N2O production as a function of 15N-labelled urea addition and multiple moisture contents (28, 40, 52, and 64% WFPS) in a Black Chernozem soil (high SOM: 55 g organic C kg−1). More N2O was sourced from SOM than added urea, with 59 ± 2% N2O originating from SOM, and SOM-derived N2O under urea was larger than that of the control, revealing a positive N2O priming triggered by urea addition (19 ± 2% of the total N2O from urea-amended soils). In summary, our findings will help to improve prediction ability and mitigation strategies for GHG emissions, particularly for N2O, from agricultural soils receiving biosolids additions.

It is projected that the World’s average temperature could rise by 0.2 oC over each of the next two decades, and by 1.1 up to 6.4 oC this century. Grazed pasture irrigation practices play a vital role in meeting the global food demand of a growing population, are keys for the efficient use of water, and conjointly, could be used to meet the climate change targets by decreasing greenhouse gas emissions levels. Grazed pasture systems are a significant source of nitrous oxide (N2O), a potent greenhouse gas and ozone-depleting-substance. In pastures, N2O emissions are driven by changes in soil inorganic-N substrate supply following urine and/or fertiliser inputs. Increasing soil moisture as a result of irrigation and rainfall events potentially increases N2O production via nitrifier-denitrification and denitrification, since these are hypoxic and anaerobic processes, respectively, and increasing soil moisture reduces oxygen (O2) supply. Net surface emissions rely on subsurface gas transfer which is controlled mainly by diffusion. Soil relative gas diffusivity (Dp/Do) is the predominant parameter that describes O2 transport in soils and has been shown to be a promising integrator of soil physical conditions to better understand and mitigate N2O emissions. The value of Dp/Do is a function of the soil air content and gaseous phase tortuosity, both of which vary with soil physical properties including soil texture, soil structure and water content. The principal outcomes of this project were: • The confirmation that Dp/Do can be used as a useful measure to predict soil physical conditions where N2O emissions occur and that recent observations of a critical diffusivity window under 0.02, and especially close to 0.006, need to be avoided when managing pastures if excessive N2O emissions are to be prevented. • Soil texture and structure were shown to have an effect on drainage rate which in turn directly affected Dp/Do, and thus N2O emissions following irrigation events. • Optimised irrigation minimised soil anaerobic conditions when compared to standard irrigation, identified through higher soil Dp/Do, which, in turn, was shown to increase plant growth and N uptake in a lysimeter study. • Lower N2O emissions were also measured under optimised irrigation in field trials, but conversely, higher emissions were observed in the lysimeter study with optimal irrigation. It was hypothesised that this may have resulted from the “Birch effect” being enhanced due to rewetting after extended drying periods (reduced in the standard irrigation) that in turn could promote hypoxia and N2O generating processes. The first laboratory experiment (Chapter III) examined the consistency of Dp/Do across different soil types to predict peaks of N2O emissions. It was hypothesised that regardless of soil type, the N2O emissions would peak at the previously reported Dp/Do value of 0.006. Increasing soil bulk density and soil matric potential caused Dp/Do to decline. As Dp/Do declined to a value of 0.006 N2O fluxes increased, peaking at a Dp/Do value of ≤ 0.006. This experiment shows that the elevation of N2O fluxes as a Dp/Do threshold of 0.006 is approached, holds across soil types. However, the variability in the magnitude of the N2O flux as Dp/Do declines is not explained by Dp/Do and is likely to be dependent on factors affecting the N2O:(N2O+N2) ratio. In Chapter IV, the effect of successive wetting-drainage cycles on both Dp/Do dynamics and associated N2O and N2 emissions in two soils; a pallic silt loam and an allophanic sandy-silt, with the later also having a higher organic matter content was determined. For both soils each wetting-drainage cycle induced N2O fluxes but with 5-fold lower fluxes in the allophanic soil. Greater aggregation and sand content in the allophanic soil generated higher porosity and Dp/Do values that were almost always greater than recognised anaerobic limits (Dp/Do < 0.02). While wetting-drainage events induce N2O emissions by altering Dp/Do and the soil aeration status, the drainage of soils, especially soils high in organic matter, may enhance O2 demand generating anaerobic zones conducive to denitrification. Low N2 emissions were potentially due to pH and high nitrate concentration effects in the pallic soil. Chapter V evaluated the potential for managing soil N2O emissions by altering irrigation timing based upon modelled Dp/Do using repacked lysimeters, sown with perennial ryegrass (Lolium perenne), that received either a standard irrigation treatment (15 mm every three days), or an optimised irrigation treatment where irrigation was applied when soil Dp/Do was 0.085 (equivalent to 50% of plant available water) and top up below Dp/Do = 0.02 (based on the previous results). Emission factors over the first 39 days of the experiment for optimised and standard irrigation treatments with urine were 1.26 and 0.12%, respectively, with an increase of N2O emissions in the optimised irrigation possibly due to the “Birch effect”. Cumulative pasture dry matter production and N uptake were higher in the optimised irrigation treatment. Macropores flow of urine derived N was lower under optimised irrigation. Finally, Chapter VI re-evaluated the effect of irrigation scheduling (standard and optimised), using soil gas diffusivity as a decision tool to mitigate N2O emissions, in the field using automatic chambers. The optimised irrigation treatment was favoured as there was a significant decrease in cumulative N2O emissions compared to the standard irrigation. Despite growth rates lower than the summer average for fertilised irrigated pastures, dry matter production for both irrigation treatments with urine were not significantly different. It was concluded that Dp/Do appears to be a relevant parameter to predict N2O emissions and future implementation into models is conceivable. Moreover, this study demonstrates the need to adjust irrigation cycles to prevent excessive drying of the soil profile between water events while anticipating the rain events and keeping Dp/Do > 0.02. Further detailed studies are needed to: (i) examine the interaction between soil structure and soil organic matter content and their effect on N2O emissions under wetting-drainage events, with measures of soil O2 (ii) consider avoiding intra-aggregate disruption through irrigation management which might minimise the Birch effect while also helping to minimise N2O emissions through reducing the increase of C supply for denitrifiers.

Soil organic carbon (SOC) is the largest terrestrial C pool and understanding SOC decomposition in response to environmental factors is critical for accurate predictions of climatic change. Soil moisture is one of the most important, yet less explored, environmental factors controlling soil microbial respiration. The relationship between soil moisture and respiration also varies with soil texture. Currently, it is difficult to predict feedbacks to climatic changes from changes in soil moisture, as most earth system models lack site-specific, experimentally-derived parameters to represent soil moisture-texture-respiration relationships. The overarching goals of this dissertation are to gain a fundamental understanding of the interactive effect of soil moisture and texture on microbial processing of SOC, and to elucidate microbial response to global changes. To accomplish this, two laboratory experiments and one field study were conducted. Laboratory experiments included (i) 90-day incubation with steady-state moisture, and (ii) 140-day incubation with alternating wetting and drying (transient) and steady state moisture using three soils of different textures (sandy, loamy, and clayey). A field experiment was conducted to understand the effect of moisture to SOC dynamics under in-situ conditions and involved drought, rainfed, and irrigated moisture treatments. The 90-day experiment revealed that texture was the major determinant of SOC cycling despite observing different moisture optima for the highest respiration for different textures. The 140-day incubation aimed to decipher mechanism(s) fueling the Birch effect (a spike in respiration rate upon rewetting of dry soils). I found that different mechanisms caused the Birch effect in different soils. In sandy soil, metabolite accumulation and changes in bacterial community, while in loamy and clayey soils, metabolite accumulation and release of aggregate protected C contributed to the Birch effect. Results imply that response of microbial respiration to changing climate will strongly depend on soil moisture sensitivity to textural differences. Field experiment results showed strong influence of moisture on SOC decomposition, i.e., lower CO2 emissions from drought plots, however, higher extractable organic C, microbial biomass C and fungal hyphal abundance were observed under drought conditions. The laboratory and field experiment provide critical information for improving the moisture sensitivity of microbial parameters in ecosystem models.

WRF-Hydro, configured as the NOAA National Water Model, is a hydrologic modeling framework that has been widely used to simulate streamflow at very fine spatial and temporal scales over the continental United States (CONUS). The initial condition of soil moisture at the beginning of the storm has a strong influence on a basin’s hydrological response, as it affects the partitioning of rainfall into runoff and infiltration. Therefore, understanding the role of initial soil moisture in rainfall-runoff modeling is important for accurately simulating streamflow and soil moisture. This study investigated the impact of initial soil moisture on the accuracy of streamflow simulations, and developed calibration schemes to improve the accuracy of streamflow and soil moisture simulations by utilizing soil moisture. Specifically, this dissertation had three main objectives: (1) quantify the sensitivity of the Weather Research and Forecasting hydrological modeling system (WRF-Hydro) to initial soil moisture conditions, (2) evaluate the sensitivity of WRF-Hydro to different sources of initial soil moisture, and (3) evaluate whether using soil moisture to calibrate WRF-Hydro can improve the model accuracy. Results from objective 1 revealed that both the spatial pattern and the value of initial soil moisture affect the accuracy of streamflow simulations. Thresholds were identified for each watershed. When soil moisture is above these thresholds, the spatial variability of initial soil moisture has a stronger influence on streamflow simulations. When the initial soil moisture is below this threshold, the spatial variability of initial soil moisture has little influence on streamflow simulations.Results from objective 2 showed that, at the watershed scale, in situ soil moisture is more accurate than SMAP downscaled soil moisture. However, the accuracy of the in situ soil moisture is influenced by the number of measurements with the watershed. Generally, the simulations are more accurate when there is a larger number of stations. Moreover, streamflow simulations are more sensitive to soil moisture during heavy precipitation events than during medium and light events. Lastly, objective 3 demonstrated that the inclusion of soil moisture into the calibration of WRF-Hydro improves the accuracy of soil moisture simulations, although it tends to degrade the accuracy of streamflow simulations. The inclusion of soil moisture in the calibration also decreased the number of iterations that were required to determine the optimal set of parameters for WRF-Hydro. Model calibration has a stronger influence on the streamflow simulations than the soil moisture simulations. Overall, this doctoral research demonstrates that incorporating in situ soil moisture in WRF-Hydro is useful for improving the accuracy of initial soil moisture and the accuracy of streamflow simulations. The results of this research can help to improve the accuracy of the National Water Model (NWM) which will result in better flood predictions and improved water management.

Diffuse phosphorus (P) loss from agricultural land contributes to surface water quality degradation. To mitigate losses and protect waterways it is important to describe all the ways that P can be mobilised and transported from land. The main aim of this thesis was to investigate P loss associated with the reductive dissolution of iron (Fe) and manganese (Mn) oxides in the soil, due to waterlogging and reducing conditions. The overall hypothesis was that anaerobic and reducing conditions in soils contribute significantly to potential annual P losses. A combination of laboratory analysis and field work was conducted in New Zealand and Ireland. The first objective was to quantify the potentially reducible phosphorus (P) component in the lab for a range of stored soil samples, and the distribution of reducible P pools relative to known soil maps. In New Zealand, eight stored soils were tested in the laboratory (Andisol, Cambisol, Vitric andisol, Ferrasol, Luvisol, Gleysol, Arenosol/Fluvisol, and Acrisol), and five stored soils were tested in the laboratory from Ireland (Cambisol, Gleysol, Luvisol, Phaeozem, and Podzol). Current models use oxic soil tests, which may not represent anaerobic conditions. Anoxic water extractable P (anoxic WEP) and sodium-bicarbonate-dithionite extractable P (dithionite-P) tests were developed to predict soil P vulnerable to reductive dissolution and potential loss under anaerobic conditions. In New Zealand and Ireland, anoxic WEP and dithionite-P varied by soil order and land use, and anoxic WEP was greater than oxic WEP¬, which showed the short-term impact of soil anoxia on P release. Models predicting anoxic WEP and dithionite-P at the 1:50,000 scale in New Zealand found relatively small proportions of agricultural land were enriched in dithionite-P (31% >85 mg kg-1) or anoxic WEP (3% >0-0.291 mg L-1). The second objective was to determine if redox reflects changes in P and redox-sensitive components, with attention to the length of time that the soil is saturated, P fertiliser treatment, and temperature. The second laboratory experiment was an incubation comparing the rate and extent of P release across contrasting soil textures (clay loam, silt loam, sandy loam), three long-term P fertiliser levels, at two temperatures (3oC v.s. 18oC), and two oxygen levels (oxygen < 0.5 mg L-1 v.s. > 7 mg L-1). The mean dissolved reactive P (DRP) concentration and its rate of release increased with fertiliser application, temperature, and in two soils, anoxic conditions - commensurate with the depletion of nitrate (NO3-) and the reductive dissolution of Fe and Mn. The release of P was complete within 24 hours, which showed that the reaction, and potential for enhanced P loss, could occur within a day of saturation. The third objective was to observe trends in the release of P and other redox-sensitive species into soil solution under saturated conditions, during the drainage season. A field experiment monitored a New Zealand soil profile with 16 unsaturated zone samplers installed down to 20 and 80 cm below ground level, from May to September of 2017 and 2019. They were installed to observe P and Fe release as the profile wets up or dries out. Events that saturated soil caused reducing conditions that released up to 77% and 96% greater P and Fe, respectively, than average over the rest of the sample period. Artificial saturation experiments in the laboratory and in the field used the same soil as the field experiment and showed that soils treated with NO3- released up to 86% less DRP and 98% less Fe than the soil that received no N. This showed that the reduction reaction was buffered by the presence of NO3-. The fourth objective was to determine the role of redox processes in the release of P and other redox species within and across drainage and runoff events. Flow-weighted artificial drainage and runoff samples were collected from four hydrologically-isolated plots on the same slope in Ireland in 2009, 2018 and 2019. They were analysed for P and redox sensitive components and assessed throughout seasons and storm events. At the site used, dissolution of P and Fe dissolution had a stronger relationship in reducing conditions. Additionally, the reaction was not season specific. The main drivers of Fe and P release were the extent of waterlogging in the soil, and levels of NO3-. In conclusion, an area that is prone to saturation excess could have increased P loss due to reductive dissolution. The research presents two updated soil test methods and predictive equations that could be used to improve estimations of P loss from poorly drained areas and specific soil types. Laboratory data gives evidence that soil texture, P fertility levels, temperature and oxygen conditions affect the amount and rate of WEP released within 24 hours of waterlogging. Field data shows that the relationship between P and Fe can be seen in soil solution as well as at the drainage output. The present work also shows that waterlogging and NO3- levels are key drivers in the occurrence of P and Fe dissolution from soil. Therefore, the research highlights periods where warm temperatures and high moisture conditions coincide as high risk. Having a high P/low NO3- system, particularly in sites that are prone to saturation events may exacerbate P release into waterways in the future. The influence of waterlogging and anaerobic conditions on Fe and P should be integrated into current critical source area models and used to inform potential mitigation strategies.

Spatial patterns of moisture and tree species have been studied using environmental gradients, often represented by terrain attributes in GIS. With climate change, GIS terrain variables, which are static as long as the elevation remains unchanged, will not reflect alterations in temperature, water cycle, and atmospheric conditions. In this thesis, the commonly used terrain variables and climate water balance variables were evaluated and compared for their ability to explain soil moisture and tree species distributions in a forested watershed in the Northern Cumberland Plateau. The results suggest that GIS terrain variables generally perform better than climate water balance variables, however, the differences are not significant for soil moisture or half of the species studied. Topographic position, a terrain attribute not explicitly considered by climate water balance variables, performed well in its ability to explain both soil moisture and tree species distributions. This suggests that the inclusion of topographic position into or in tandem with future iterations of climate water balance variables could be advantageous.

With a growing population and climate-related disasters projected to become more severe, the need to anticipate and better prepare for food crises has never been higher. Status quo early warning systems perform well, by and large, but complex weather phenomena are a major source of uncertainty. In this dissertation, I explore whether statistical diagnostics associated with tipping point theory can be used to improve detection of major crises.First, the accuracy of early warning systems is evaluated – with a focus on the Greater Horn of Africa. I discuss geographical disparities in early warning skill, and explore the various sources of uncertainty that impact food security projection accuracy. I quantify the contribution of climate forecast skill and conflict events in overall early warning skill to use as a baseline upon which to improve.Second, the potential applications of tipping point theory are discussed in relation to the use of remotely-sensed environmental variables for food security early warning. I evaluate the different statistical diagnostics that can be applied to identify food security tipping points and the data requirements in terms of spatial and temporal resolutions, data availability, latency and geographic coverage in order to be maximally useful for early warning analysis. Finally, I test whether statistical diagnostics of tipping points can be used for detection of food security transitions. Through this work, I demonstrate the utility of combining soil moisture measurements from the SMAP mission with food price data for identifying transitions, with a lead time of approximately 3 months. The model presented here was used to detect all major food security transitions (both towards and away from a crisis) in the observational record of soil moisture. Importantly, the diagnostics are also correlated with the magnitude of change in food security conditions which is critical information for food security planning. Overall, the dissertation advances the science and capabilities of food security early warning, which will lead to an improvement in saving lives and resources.

This study aims to evaluate the capability and transferability of a physically-based hydrologic model to understand the trade-offs between precipitation, soil moisture and surface energy fluxes at sites with different vegetation types in the U.S. Southern Plains. One of the benefits of training a process-based model is the capacity to use it as a complement to standard weather stations for predicting energy fluxes, soil temperature and moisture estimations. Modeling of the terrestrial surface soil moisture and temperature, and boundary layer energy fluxes is key for understanding the spatio-temporal variability of hydro-meteorological conditions that drive normal and extreme (i.e. floods and droughts) events. Soil moisture (SM), surface energy fluxes (SEF) and soil temperatures (ST) play an important role in the ground and near surface hydro-energetic dynamics, especially in water exchange processes such as the evapotranspiration (ET). ET is an important variable for understanding the energy, water and biogeochemical budgets. This study uses the Triangulated Irregular Network TIN-based Real Time Integrated Basin Simulator (tRIBS), a continuous physically-based distributed hydrological model, to provide estimations of the surface energy balance (SEB) components in typical environments of the U.S. Southern Plains. Both calibration and validation of the model are performed using available Eddy Covariance Tower (ECT) observations distributed on crops and grasslands in Oklahoma. The model calibration is based on a hybrid strategy that uses a manual procedure followed by an optimization algorithm based on the Shuffled Complex Evolution (SCE) theory. All data used to parametrize the model is free-access. Satellite data is needed to represent dynamic vegetation conditions of albedo, leaf area index, vegetation throughfall coefficient, stomatal resistance and vegetation fraction. Model calibration is conducted during one hydrologic year at two stations with differing vegetation cover. Model validation is conducted at the same ECTs during a different year than the calibration. Transferability of the model parameterization is tested at other ECT with similar vegetation conditions in the Southern Plains. The model calibration and validation results showed the strong capabilities of tRIBS to predict the energy fluxes (Nash>0.5) and the soil temperature profile (Nash>0.7). In addition, the model predicts the soil water content well, but snow occurrences affected the model’s performance. Furthermore, tRIBS captures the seasonal and diurnal cycles of the energy partitioning with the use of remotely-sensed dynamic vegetation from MODIS and standard weather observations. Vegetation changes play a crucial role in the simulations accuracy. The onset of vegetation greening impacts the contribution of the different components of ET with increased contribution of plant transpiration (T) over soil evaporation (E). Finally, the model has potential to be transferable and of use as a forecasting tool in similar environments in the Southern Plains without parameter calibration. However, the quality of the weather forcing and satellite data will influence the accuracy of the results.

Abstract: Snowpack accumulation and melt dynamics, and soil moisture storage (SMS) play a critical role in regulating stressed regional water supplies in the southern Alberta Rockies. However, these processes are sensitive to impacts from forest cover losses by timber harvesting and associated land disturbance. While numerous studies have investigated effects of clearcut harvesting, few studies have evaluated the comparative effects of alternative forest-harvesting methods on spatial and temporal snowpack accumulation-melt and subsequent growing season soil moisture dynamics, resulting in a knowledge gap about how current alternative or new/novel strategies influence these key hydrological processes. Spatio-temporal patterns of snowpack accumulation/melt and SMS were assessed from 2016-2018 in the three sub-watersheds located in the Crowsnest Pass, Alberta that were harvested using three different forest-harvesting strategies (clear-cut with tree retention, strip-shelterwood, and partial cut harvesting). Snowpack and SMS were measured in each harvest and adjacent undisturbed forest stands using a series of linear transects. Snow depth and water equivalent (SWE) was measured every 14-21 days from the period of peak snowpack through snow disappearance, and SMS monitored at the same frequency thereafter using time domain reflectometry (TDR) to measure soil water storage at 0-20 cm and 20-60 cm depths. Differences in snowpack accumulation and melt and SMS were evaluated in the context of measured differences in microclimate and modeled solar radiation forcing. Results from this study showed peak snow accumulation increased by 43% in harvested strips, and 28% in the partial-cut relative to corresponding forested reserves, while snowpacks in north and south facing clearcut were either not different or significantly less. Differences in solar radiation due to slope orientation and shading by retained trees produced highly varied patterns of snow accumulation and melt creating within-harvest differences in snow disappearance timing of up to 20 days. Subsequent seasonal SMS was also generally increased by harvesting due to reduced water losses from tree transpiration. However, the duration and magnitude of differences in SMS varied among harvests and soil layer depths over wet and dry years, which was attributed to differences in the ability of specific harvest cutting-patterns to moderate surface soil moisture losses. Across both study years, increased SMS after harvesting was greatest in SS (39%) followed by CC (~25% across both north- and south-facing hillslopes), and PC (~19%) for the two soil layers considered (0-20, and 20-60 cm). Results from this study demonstrate key differences in the comparative ability of specific patterns of tree retention to moderate spatial patterns of snow accumulation and melt and seasonal SMS, and strongly reinforce past research on the importance of interactions between forest cover, slope-orientation, and solar radiation in controlling spatial and temporal water-balance. These findings provide new information on relationships between alternative forest harvesting methods and snowpack and SMS dynamics needed to help evaluate integrated forest-water management strategies in this vital water supply region.

Constructing athletic fields on a sand base allows for more rapid drainage; however, lateral shear strength and traffic tolerance tends to not be as good as that of a native soil field. Hybrid turf systems were developed that intertwine turfgrass with synthetic fibers that are able to improve shear strength. Hybrid systems are used on sports fields ranging from practice fields to competition fields globally. Previous stabilization systems included fibers that were shorter than the intended height of cut of the turfgrass. The purpose of this study was to test a new system which has the natural turf maintained at the same height as the synthetic fibers. The project’s main goal was to compare two hybrid turf systems constructed in combination with two different types of sand (round vs angular), and three turf species to determine which leads to the highest performance for elite level soccer matches, while also keeping player safety in mind. Measurements and treatments taken to achieve these goals included Gmax, traffic, quality, soil moisture and shear strength both rotational and lateral. All measurements and data were taken at the project’s field site at the Ohio Turfgrass Foundation (OTF) research facility over the course of two years. Results indicated that turf species was ultimately the most significant factor, with both hybrid carpets doing their intended job and sand shape playing a much smaller roll than initially anticipated on overall results. The HERO hybrid system in conjunction with angular sand and ryegrass was the overall best combination for elite level soccer teams to be able to have both safer fields and higher levels of player performance.

Forests have the ability to sequester carbon from our atmosphere. Soil respiration (Rs) plays a role in a forest's ability to do so as it is a significant source of carbon dioxide back to the atmosphere. Therefore, understanding the process of Rs under varying conditions is gaining more attention. As of now we have a relatively good understanding of Rs under managed forest ecosystems such as pine plantations. This particular study examined Rs under different overstories and understories in a high elevation Southern Appalachian forest in order to get a better understanding of Rs under a natural hardwood system. The four vegetation types under consideration were an eastern hemlock (Tsuga canadensis L. Carriere) dominated overstory, a hardwood overstory with little to no understory, a mountain laurel (Kalmia latifolia L.) dominated understory, and a cinnamon fern (Osmundastrum cinnamomeum (L.) C.Presl) dominated understory. Differing temporal variations of Rs were observed under the vegetation types. We found monthly differences in rates among vegetation type however, an overall annual difference in Rs rates between vegetation types was not observed. This simply indicates the importance of observing Rs under different time scales to get a better understanding of its variation. We also calculated vegetation indices from remotely-sensed data to explore any relationships to Rs as well as if the indices themselves could improve out model. A vegetation index is a number that is calculated for every pixel in a remotely sensed image and represents plant vigor or abundance. Few significant relationships were found between the indices and Rs. Future work may want to better understand vegetation indices' spatial extent and accuracy in order to find whether they may be beneficial in Rs estimation. Understanding the influence of varying vegetation type and soil temperature and moisture on Rs will ultimately improve our ability to predict what drives changes in carbon fluxes.

Accurate soil moisture measurement plays a crucial role in optimum irrigation management, quantification of crop water use and resolution of soil water balance, besides numerous other applications. Numerous sensors exist to estimate soil moisture and due to their diverse principles of operation, they are subject to varying levels of errors and uncertainties. Thus, prior to their use, the sensors have to be comprehensively assessed for their performance so as to evaluate the degree and nature of inaccuracy as well as the limits to their applications. In this research, nine commercially available soil moisture sensors (CS655, CS616, 10HS, 5TE, EC-5, Teros 21 (MPS-6), SM150, TDR315L (Acclima) and JD Probe) were evaluated in silt loam and loamy sand soils. These sensors were investigated for their performance under two different orientations, i.e., vertical and horizontal. These investigations were conducted for two growing seasons (2017 and 2018) at South Central Agricultural Laboratory (an irrigated row crop site with maize-soybean rotation) and Central City (a rainfed pasture). In the first chapter of this thesis, performance statistics for various sensors were quantified under factory calibrations (F.C.) during 2017. To account for the uncertainty associated with estimating volumetric water content (θv), calibration functions were developed and were validated using independent datasets collected during 2018. In the second chapter, the suitability of the sensors to report (a) total soil water (TSW); (b) actual evapotranspiration (ETc); and (c) irrigation triggers (IT) was assessed, which is rarely addressed in the literature. Finally, in the third chapter, we proposed a step-by-step guide that aimed at appropriate sensor selection by users with widely varying expectations and conditions of use. Each sensor was evaluated based on its operational feasibility that included cost conditions, telemetry availability, and ease of operation, and its performance accuracy. This research is unprecedented, to the best of our knowledge, and aims at a comprehensive evaluation of sensors for factors that are not commonly recognized when assessing soil moisture sensing technology. Immense potential exists for the application and use of the findings of this research for sensor selection by diverse users is concerned. Advisor: Suat Irmak

Water movement through the landscape to streams provides a fundamental linkage between terrestrial and aquatic environments in headwater systems. Headwater streams, which are the smallest and most abundant streams, are critical components of drainage systems, connecting important terrestrial and aquatic biogeochemical cycles and influencing the nutrient dynamics of downstream ecosystems. Given the importance of water presence and movement as a driver of biogeochemical transformations and the transport of material from terrestrial to aquatic ecosystems, the primary goal of this research was to quantify the distribution across and movement of water and other elements through a forested watershed to a headwater stream and ultimately to an inland lake in the glacial drift landscape of northern Michigan, U.S.A. To investigate the temporal dynamics and spatial patterns of water across the watershed and in Honeysuckle Creek, stream discharge, shallow groundwater levels, soil moisture, and water chemistry were measured from 2015–2017. Along the stream, surface flow was seasonal in the main stem and perennial flow was spatially discontinuous for all but the lowest reaches. Spring snowmelt was the dominant hydrological event in the year with peak flows an order of magnitude larger than annual mean discharge. Topography and soil characteristics strongly influence water and dissolved matter movement through the landscape. Water presence across the watershed was highly variable with perennial soil saturation and shallow groundwater within 10 cm of the surface at the lowest landscape positions, low soil moisture and nonexistent groundwater in upland outwash ecosystems, and a mixture of these conditions in the heterogeneous till parent material ecosystems of middle landscape positions. Shallow groundwater was the primary source of water to the headwater stream throughout the year; originating from snowmelt in the spring and shifting to recent precipitation in summer and autumn seasons. The riparian areas and the outwash-lake plain wetland likely have a much stronger influence on stream chemistry and discharge than the upland landscape units, due to the perennial connections between shallow groundwater from these wetland areas and surface water.

Drought is one of the most severe natural disasters and detrimentally impacts water resources, agricultural production, the environment, and the economy. Climate change is expected to influence the frequency and severity of extreme droughts. This dissertation evaluates drought conditions using a variety of hydrologic modeling approaches include short-term drought forecasting, long-term drought projection, and a coupled surface-groundwater dynamic drought assessment. The economic impacts of drought are also explored through a linked economic impact model. Study results highlight the need for various drought assessment approaches and provide insights into the array of tools and techniques that can be employed to generate decision-support tools for drought mitigation plans and water resource allocation. For short-term drought forecasting, the Soil and Water Assessment Tool (SWAT) and Variable Infiltration Capacity (VIC) models are used with a meteorological forecasting dataset. Results indicate that eight weeks of lead-time drought forecasting show good drought predictability for the Contiguous United States (CONUS). For the drought projection at a finer scale, both SWAT and VIC models are applied with Coupled Model Intercomparison Project Phase 5 (CMIP5) climate model outputs to derive multiple drought indices for the Chesapeake Bay watershed and five river basins in Virginia. The results indicate that current climate change projections will lead to increased drought in the entire Chesapeake Bay watershed and Virginia river basins because of increases in the sum of evapotranspiration, and surface and groundwater discharge. The impacts of climate change on future agricultural droughts and associated economy-wide implications are then evaluated using the VIC and IMPLAN (IMpact analysis for PLANning) model for the several congressional districts in Virginia. The result indicated that increases in agricultural drought in the future would lead to decreases in agricultural productions and job losses. Finally, a coupled framework using the VIC and MODFLOW models is implemented for the Chesapeake Bay and the Northern Atlantic Coastal Plain aquifer system, and the results of a drought index that incorporates groundwater conditions performs better for some drought periods. Hydrologic modeling framework with multiple hydrologic models and various scales can provide a better understanding of drought assessments because the comparisons and contrasts of diverse methods are available.

Agronomic and physiological performance of four perennial grasses under summer dry conditions in New Zealand The main aim of this research was to quantify the effects of different levels of moisture and nitrogen (N) supply on dryland brome, cocksfoot, perennial ryegrass and tall fescue pastures grown under summer dry conditions. This was assessed from establishment (2014/15) to the end of the second year (2015/16) in two field experiments. Moisture and temperature are the major factors that affect germination. To quantify their combined effects, a laboratory experiment was used to calculate hydrothermal time requirements. Increasing temperature from 25 to 30 and 35 °C caused a rapid decline in germination rate of cocksfoot and tall fescue even when moisture was not limited (0 MPa). Applying moisture stress, especially when temperature was higher than 20 °C resulted in a more rapid decline in cocksfoot germination rate and final germination percentage than in perennial ryegrass, tall fescue and brome. To explore the effects of different soil moisture levels on pasture production in the field, two experiments were established in October 2014 at Ladbrooks and Ashley Dene, which differed in plant available water content (PAWC). The Wakanui soil at Ladbrooks has a 41-55% higher PAWC than the stony Lismore silt soil at Ashley Dene. A Latin square design was established at both sites with four grass species and four replicates. During establishment period, from October 2014 to June 2015, total dry matter (DM) production at Ladbrooks was 4.34 t ha-1 compared with 1.93 t ha-1 at Ashley Dene, with no differences among species. In the second year from July 2015 to July 2016, the experiments included a nitrogen (N) treatment. At Ladbrooks 900 kg N ha-1 and at Ashley Dene 500 kg N ha-1 were applied. In 2015/16, mean annual DM production was 19.8 t ha-1 for +N and 9.20 t ha-1 for −N treatments at Ladbrooks. At this location, there were no differences for DM production among the species. In 2015/16 at Ashley Dene, the +N cocksfoot treatment DM was 6.0 t ha-1 higher than the average of 3.2 t ha-1 for the other ±N species. This difference at Ashley Dene was explained by the interaction between species and N after 85 mm of rainfall in mid-January of 2016, when only the cocksfoot responded to the applied N. The impact of seasonal temperatures on DM yield was quantified using thermal time accumulated above a base temperature of 3 °C. In the 2015/16 period, the relationship between accumulated thermal time and DM was separated into two phases at Ladbrooks and into five phases at Ashley Dene. These phases indicated decreased or increased growth due to the onset of moisture stress or rainfall. At Ladbrooks, the break points were not different among species, but were later in +N compared with –N treatments. For +N pastures linear growth in Phase 1 lasted until 6/3/2016, after 2355 ± 44.6 °Cd. For −N treatments Phase 1 stopped after only 2161 °Cd, on 19/2/2016. The temperature adjusted linear growth rate was differed among species and N treatments. For cocksfoot and perennial ryegrass it was 6.94 kg DM °Cd-1 ha-1 which was higher than brome and tall fescue (6.00 kg DM °Cd-1 ha-1). At Ashley Dene, the additional yield from +N cocksfoot resulted in a higher linear growth rate in Phase 3, after the 85 mm of rainfall in January 2016. Plant available water content was 254 mm for +N and 195 mm for −N pastures at Ladbrooks compared with a mean of 101 mm at Ashley Dene. In the 2015/16 period, mean total water use was 823 mm for +N pastures which was higher than the −N pastures (777 mm) at Ladbrooks. There were no differences among species. Mean water use efficiency for +N pastures was 20.4 kg ha-1mm-1 which was higher than −N pastures (8.36 kg ha-1mm-1). There were no differences among the species. In 2015/16, total water use at Ashley Dene was 364 mm and was not different among species. However, water use efficiency for cocksfoot was 16.6 kg ha-1mm-1 which was higher than other species. The higher production of + pastures at Ladbrooks was explained by higher radiation interception and also radiation use efficiency (RUE). The lower yield at Ashley Dene compared with Ladbrooks was due to the reduction in the amount of photosynthetically active radiation intercepted by the canopies as well as a lower RUE. Higher radiation interception of +N cocksfoot on 20/1/2016 and 21/4/2016 was the main explanation for its higher DM production. Overall, the range of dry matter production yield from the different soil moisture and N conditions could be quantified by calculation of growth rates in relation to thermal time, soil water extraction patterns, estimation of light interception and radiation use efficiency, explain the physiological basis of yield differences.

This research compares Gravity Recovery and Climate Experiment (GRACE) groundwater storage (GWS) and root zone soil moisture (RZSM) percentiles to measured data, other drought indicators (DIs) and indices, and stakeholder observations for the purpose of assessing the feasibility and usefulness of these products to detect drought conditions. GRACE percentiles were directly compared to historic groundwater percentiles at 89 Nebraska well locations. Spatial time-series correlations over CONUS were performed between GRACE GWS and RZSM and the U.S. Drought Monitor (USDM), Standardized Precipitation Index (SPI), and soil moisture parameters from several North American Land Data Assimilation System (NLDAS) models. A survey of stakeholder observations during a 2016 flash drought event centered on Montana, Wyoming, South Dakota, and Nebraska was also compared to GRACE percentile data to analyze drought onset timing, geographic coverage, and severity. Overall the results show GRACE GWS has similar spatial and temporal agreement over the well period of record, and generally has the expected negative correlation relationship with observed groundwater, but it does not accurately reflect historic percentiles in Nebraska. GRACE GWS and RZSM have moderate correlation with USDM, and high correlation with SPI, and NLDAS models over the entire U.S. with notable regional and seasonal patterns. SPI accumulation period also plays an important role in correlation strength for both RZSM and GWS with the best agreement seen at 3-month and 12-month accumulation periods, respectively. GRACE RZSM time-series data closely matches stakeholder observations of decreasing soil moisture availability, while observations of decreasing water levels were not as closely matched by GWS. When analyzed as an average over all responding zip codes, RZSM showed an early warning trend up to six weeks prior to observed reports. These results indicate GRACE percentiles are promising drought indicators that can be used as a monitoring and early warning system by decision makers. Advisor: Tsegaye Tadesse

Water is one of the most precious natural resources, and sustainable water resources development ‎‎is a significant challenge facing water managers over the coming decades. Accurate estimation of ‎‎the different components of the hydrologic cycle is key for water managers and planners in order ‎‎to achieve sustainable water resources development. The primary goal of this dissertation was to ‎investigate techniques to combine datasets acquired by remote and proximal sensing and in-situ ‎sensors for the improvement of monitoring near surface water fluxes. This dissertation is ‎separated into three site-specific case studies. First study, investigated the feasibility of using ‎inverse vadose zone modeling for field actual evapotranspiration (ETa) estimation. Results show ‎reasonable estimates of ETa, both daily and annually, from soil water content (SWC) sensors and ‎Cosmic-Ray Neutron Probes (CRNPs). Second study, combined remote and proximal sensing ‎methods to explore the spatial correlation between hydrological state variables and ET flux. ‎Comparison of the datasets reveal that SWC and ETa were linearly correlated but the correlation ‎between depth to the water table and ETa was weak. A simple multivariate linear regression ‎model was used to estimate ETa. The estimated ETa values were then compared to the time ETa ‎integration spline method. The comparison indicates similar seasonal ETa between the two ‎methods in 2015 ‎‎(wet) but a 20% reduction in 2016 (dry). The study highlights the challenge of ‎connecting hydrologic state variables with hydrologic flux estimates. Third study, evaluated the ‎functionality of automatically calibrated Earth Engine Evapotranspiration Flux (EEFlux) to the ‎existing mapping evapotranspiration at high resolution with internalized calibration (METRIC) ‎images in different locations. The comparison results showed that EEFlux is able to calculate ‎Reference evapotranspiration Fraction (ETrF) and ETa in agricultural areas comparable ‎‎(RMSE=0.13) to the ones from trained expert METRIC users. However, the EEFlux algorithm ‎needs to be improved to calculate ETrF and ETa in non-agricultural areas (RMSE=0.21). Given ‎the paucity of in-situ data across much of the globe the field of remote sensing offers an ‎alternative but requires users to be cautious and realistic about associated errors and uncertainty ‎on using such information to help construct a hydrologic budget.‎ Advisor: Trenton E. Franz and Ayse Kilic

In urban areas, increased runoff from storm events is a significant concern due to flooding, erosion, ecosystem disturbance, and water quality problems. Phosphorus (P) and nitrogen (N) are key limiting nutrients that lead to eutrophication and harmful algal blooms in Lake Erie and other freshwater systems, and urban landscapes are locally important sources of nutrients to downstream water bodies. Green infrastructure (GI) is one increasingly popular solution being used in urban areas to address the water quantity and quality issues that urban runoff creates. Green roofs and bioretention cells are widely used forms of GI designed to decrease and slow down runoff through evapotranspiration and infiltration. I compared the hydrologic effectiveness of a co-located extensive green roof and two bioretention cells in northeastern Ohio, in order to understand their relative capacities to decrease and slow down stormwater runoff, when subjected to the same weather conditions. I also monitored the green roof to understand its effects on water quality, in terms of P and N.From June 2015 to November 2016, 93 storms from 2.5 to 62 mm, were monitored. To assess the hydrologic performance of each site, I measured rainfall, underdrained outflow, groundwater levels in the bioretention cells, and soil moisture on 1–5 minute intervals. To assess water quality for the green roof, I collected precipitation and samples from the green roof downspout. I measured chloride (Cl-) phosphate (PO43-), total P, nitrate (NO3-), ammonium (NH4+), and total N concentrations using ion chromatography and a colorimetric assay.The bioretention cells performed similarly to each other, despite slightly different designs, and they had superior performance to the green roof. The paved lot bioretention cell showed 77% volumetric and 85% peak flow reduction and the gravel lot bioretention cell showed 78% volumetric and 82% peak flow reduction. The green roof only reduced 59% of its water input and 69% of peak flow. Overall, precipitation amount and peak rain intensity were the most significant predictors of hydrologic performance.The green roof exported PO43- with an average storm event mass balance of -20,115% and retained NO3- with an average mass balance of +77%. As much as half of the total P outflow concentrations consisted of PO43-. However, outflow PO43- concentrations did decrease by ~3x after fertilizer application ceased in 2016. Many cities are aggressively promoting green roofs as part of their stormwater management policies. If these policies result in many green roofs that are directly connected to downstream water bodies, phosphorus export may exacerbate water quality problems in urban areas. Re-consideration of the maintenance and construction of green roofs may be required before further adoption of the practice.

Soil moisture information can be used accurately in determining the timing and amount of irrigation applied to plants. Pan and Pan et al. proposed a robust and simple daily diagnostic equation for estimating daily soil moisture. The diagnostic equation evaluates the relationship between the soil moisture loss function and the summation weighted average of precipitation. The loss function uses the sinusoidal wave function which employs day of the year (DOY) to evaluate the seasonal variation in soil moisture loss for a given year. This was incorporated into the daily diagnostic equation to estimate the daily soil moisture for a location. Solar radiation is an energy source that drives the energy and water exchanges between vegetation and the atmosphere (i.e., evapotranspiration), and thus impacts the soil moisture dry-down. In this paper, two parameters (the actual solar radiation and the clear sky solar radiation) are introduced into loss function coefficient to improve the estimation of soil moisture. After the Introduction of the solar radiation data into soil moisture loss function, a slight improvement was observed in the estimated daily soil moisture. Pan observed that generally the correlation coefficient between the estimated and the observed soil moisture is above 0.75 and the root mean square error is below 5.0 (%v/v). The introduction solar radiation data (i.e. clear sky solar radiation and actual solar) improve the correlation coefficient average for all the sites evaluated by 0.03 when the root mean square error is generally below 4.5(%v/v) for the entire root zone.

Software as a service (SaaS) is a primary working pattern and a significant application model for next generation Internet application. Web GIS services are the new generation of the Software as a service that can provide the hosted spatial data and GIS functionalities to the practical customized applications. This study focused on developing a webGIS based application, Soil Moisture-Based Irrigation Scheduling Tool (SMIST), for predicting soil moisture in the next seven days using the soil moisture diagnostic equation (SMDE) and the upcoming seven precipitation forecasts made by the National Weather Service (NWS), and ultimately producing an accurate irrigation schedule based on the predicted soil moisture. The SMIST is expected to be capable of improving the irrigation efficiency to protect groundwater resources in the Texas High Plains and reducing the cost of energy for pumping groundwater for irrigation, as an essential public concern in this area. The SMIST comprised an integration of web-based programs, a Hydrometeorological model, GIS, and geodatabase. It integrates two main web systems, the soil moisture estimating web application for irrigation scheduling based on the soil moisture diagnostic equation (SMDE), and an agricultural field delineation webGIS application to prepare input data and the model parameters. The SMIST takes advantage of the latest historical and forecasted precipitation data to predict soil moisture in the user-specified agricultural field(s). In this regard, the next seven days soil moisture versus the soil moisture threshold for normal growth would be presented in the result page of the SMIST to help users to adjust irrigation rate and sequence.

Landslide characterization and hazard assessments require multidisciplinary approaches that connect geologic processes with geotechnical parameters. Field monitoring of hydrologic variables such as water content and water potential, coupled with geoelectrical measurements that can establish relationships used for geotechnical and landslide hazard investigations is deficient. This study brings together different techniques to develop a methodology that connects geoelectrical measurements and shear strength. A field-based framework was established that includes (1) analysis of long-term soil moisture fluctuations within different landslides (2) establishment of constitutive and new equations that test the use of electrical conductivity to predict soil-water relationships and shear strength (3) using electrical resistivity tomography (ERT) to support and facilitate the prediction of shear strength in a slope. Hydrologic conditions including volumetric water content, water potential, and electrical conductivity in the soil were measured at three active landslides in Kentucky. The in-situ electrical conductivity used within the framework is valid as a predictor of suction stress and shear strength. The ERT supports interpretations of landslide failure zones, landslide type, lithologic boundaries, and changes in moisture conditions, but also is able to utilize the methodology to calculate shear strength, and provide a spatial view of shear strength in the slope. The practical application of this framework is to support landslide hazard assessment and further understand the long-term influence of moisture conditions in hillslope soils. These parameters are pertinent to investigating the stability of landslides that are often triggered or reactivated by rainfall.

The ability to quantify soil moisture spatial variability and its temporal dynamics over entire fields through direct soil observations using remote sensing will improve early detection of water stress before crop physiological or economic damage has occurred, and it will contribute to the identification of zones within a field in which soil water is depleted faster than in other zones of a field. The overarching objective of this research is to develop tools and methods for remotely estimating soil moisture variability in agricultural crop production. Index-based and machine learning methods were deployed for processing hyperspectral data collected from moisture-controlled samples. In the first of five studies described in this dissertation, the feasibility of using “low-cost” index-based multispectral reflectance sensing for remotely delineating soil moisture content from direct soil and crop residue measurements using down-sampled spectral data were determined. The relative reflectance from soil and wheat stalk residue were measured using visible and near-infrared spectrometers. The optimal pair of wavelengths was chosen using a script to create an index for estimating soil and wheat stalk residue moisture levels. Wavelengths were selected to maximize the slope of the linear index function (i.e., sensitivity to moisture) and either maximize the coefficient of determination (R2) or minimize the root mean squared error (RMSE) of the index. Results showed that wavelengths centered near 1300 nm and 1500 nm, within the range of 400 to 1700 nm, produced the best index for individual samples; however, this index worked poorly on estimating stalk residue moisture. In the second of five studies, 20 machine learning algorithms were applied to full spectral datasets for moisture prediction and comparing them to the index-based method from the previous objective. Cubic support vector machine (SVM) and ensemble bagged trees methods produced the highest composite prediction accuracies of 96% and 93% for silt-loam soil samples, and 86% and 93% for wheat stalk residue samples, respectively. Prediction accuracy using the index-based method was 86% for silt-loam soil and 30% for wheat stalk residue. In the third study, a spectral measurement platform capable of being deployed on a UAS was developed for future use in quantifying and delineating moisture zones within agricultural landscapes. A series of portable spectrometers covering ultraviolet (UV), visible (VIS), and near-infrared (NIR) wavelengths were instrumented using a Raspberry Pi embedded computer that was programmed to interface with the UAS autopilot for autonomous reflectance data acquisition. A similar ground-based system was developed to keep track of ambient light during reflectance target measurement. The systems were tested under varying ambient light conditions during the 2017 Great American Eclipse. In the fourth study, the data acquisition system from the third study was deployed for recognizing different targets in the grayscale range using machine learning methods and under ambient light conditions. In this study, a dynamic method was applied to update integration time on spectrometers to optimize sensitivity of the instruments. It was found that by adjusting the integration time on each spectrometer such that a maximum intensity across all wavelengths was reached, the targets could be recognized simply based on the reflectance measurements with no need of a separate ambient light measurement. Finally, in the fifth study, the same data acquisition system and variable integration time method were used for estimating soil moisture under ambient light condition. Among 22 machine learning algorithms, linear and quadratic discriminant analysis achieved the maximum prediction accuracy. A UAS-deployable hyperspectral data acquisition system containing three portable spectrometers and an embedded computer was developed to classify moisture content from spectral data. Partial least squares regression and machine learning algorithms were shown to be effective to generate predictive models for classifying soil moisture.

Evapotranspiration is one of the largest fluxes in the terrestrial water cycle, and also impacts the surface energy budget and the carbon cycle. In this dissertation, I explore how the surface energy budget contributes to ET uncertainty, I apply new satellite soil moisture observations to improve global ET estimates, and I link the carbon and water cycles from space to characterize how vegetation responds to stressful conditions.First, the differences of ground heat flux models are evaluated against 88 locations with in situ observations. I discuss the mechanisms which control ground heat flux and quantify how errors in this measurement have the potential to impact evapotranspiration. A new optimized model is presented to reduce this potential uncertainty.Second, I apply integrated observations of soil moisture and evapotranspiration to improve a satellite-based evapotranspiration algorithm. I demonstrate how observations from soil moisture improve evapotranspiration estimates in water-limited regions and use this new model with soil moisture observations from the Soil Moisture Active Passive (SMAP) mission to compute evapotranspiration globally. I compare the new model with the original model and quantify how evapotranspiration is partitioned. This is the first global satellite-derived evapotranspiration dataset to incorporate water availability limitations from SMAP.Finally, I link independent measurements of the carbon and water cycle from space. I use satellite derived transpiration and new observations of solar induced fluorescence from the Orbiting Carbon Observatory -2 (OCO-2) to characterize how vegetation responds to hotter and drier climate perturbations. Overall, this dissertation advances remote sensing evapotranspiration algorithms through quantifying the uncertainty contribution from ground heat flux models and provides a new relationship to link soil moisture observations to evapotranspiration. Additionally, I present the first study to apply transpiration and solar induced fluorescence from OCO-2 to explore how vegetation responds to hotter and drier conditions by regulating the carbon and water cycles. This dissertation delivers new ideas of how to leverage earth observing satellites to advance ET science and address knowledge gaps in the earth system.

Few current modeling tools are designed to predict short-term, high-risk runoff from critical source areas (CSAs) in watersheds which are significant sources of non point source (NPS) pollution. This study couples the Soil and Water Assessment Tool-Variable Source Area (SWAT-VSA) model with the Climate Forecast System Reanalysis (CFSR) model and the Global Forecast System (GFS) model short-term weather forecast, to develop a CSA prediction tool designed to assist producers, landowners, and planners in identifying high-risk areas generating storm runoff and pollution. Short-term predictions for streamflow, runoff probability, and soil moisture levels were estimated in the South Fork of the Shenandoah river watershed in Virginia. In order to allow land managers access to the CSA predictions a free and open source software based web was developed. The forecast system consists of three primary components; (1) the model, which preprocesses the necessary hydrologic forcings, runs the watershed model, and outputs spatially distributed VSA forecasts; (2) a data management structure, which converts high resolution rasters into overlay web map tiles; and (3) the user interface component, a web page that allows the user, to interact with the processed output. The resulting framework satisfied most design requirements with free and open source software and scored better than similar tools in usability metrics. One of the potential problems is that the CSA model, utilizing physically based modeling techniques requires significant computational time to execute and process. Thus, as an alternative, a deep learning (DL) model was developed and trained on the process based model output. The DL model resulted in a 9% increase in predictive power compared to the physically based model and a ten-fold decrease in run time. Additionally, DL interpretation methods applicable beyond this study are described including hidden layer visualization and equation extractions describing a quantifiable amount of variance in hidden layer values. Finally, a large-scale analysis of soil phosphorus (P) levels was conducted in the Chesapeake Bay watershed, a current location of several short-term forecast tools. Based on Bayesian inference methodologies, 31 years of soil P history at the county scale were estimated, with the associated uncertainty for each estimate. These data will assist in the planning and implantation of short term forecast tools with P management goals. The short term modeling and communication tools developed in this work contribute to filling a gap in scientific tools aimed at improving water quality through informing land manager's decisions.

The recently developed Cosmic-ray Neutron Probe (CRNP) for estimating soil water content (SWC) fills a critical measurement gap between point scale methods and large scale measurements collected from remote sensing. CRNP works by measuring the change in low-energy neutron intensity over time. However, the accuracy of CRNP to measure SWC is well known to be affected by other hydrogen sources (e.g. soil organic content, atmospheric water vapor, vegetation and surface water). This study focuses on the influence of rapidly growing vegetation in agricultural fields on the accuracy of the CRNP method. Here we use data from three long-term CRNP study sites in central Nebraska (Paulman Farms, est. 2015), eastern Nebraska (US-Ne3, est. 2011), and central Iowa (IVS, est. 2010) that span a natural precipitation gradient of increasing precipitation from west to east. All three fields grow maize and soybean depending on rotation. At each CRNP site both hourly moderated and bare neutron counts are recorded. Previous research has shown that the bare to moderated ratio may be a good indicator of changing vegetation conditions and useful as a correction to estimating SWC. In addition, I use the MODIS remote sensing dataset to calculate a widely used index to monitor vegetation, Green Wide Dynamic Range Vegetation Index (GrWDRVI or WDRVI). Finally, observed vegetation data from US-Ne3 was collected biweekly from 2003-2016 and used as a benchmark for the CRNP and remote sensing analyses. My results indicate that biomass data determined from remote sensing (GrWDRVI) closely follows in-situ sampling of biomass (R2=0.677 for Maize and R2=0.567 for Soybean). The driest site (Paulman Farms) showed the best relationship between bare to moderated (B/M) ratios and GrWDRVI with an R2 = 0.9188, while the wettest site (IVS) showed the worst relationship with R2 = 0.09. I found that local correction factors using B/M ratio and moderated counts removing the influence of vegetation changes can be derived, thus removing bias in the CRNP SWC observations. The improved algorithm for estimating SWC from CRNP will be beneficial for long term monitoring as well as validating remote sensing SWC products. More experiments with direct biomass observations and needed to fully understand the relationship between GrWDRVI, bare and moderated neutron counts, and in-situ biomass. Advisor: Trenton E. Franz

Cocksfoot (Dactylis glomerata) is a preferred grass option for dryland pastures on high country farms in the South Island of New Zealand. However, increasing the productivity of cocksfoot based pastures with legumes is difficult due to the harsh environment and difficulty to apply fertiliser. Russell lupin, or perennial lupin (Lupinus polyphyllus) has been identified as a potential legume option due to its ability to tolerate low fertile and acidic soils, cold winters and hot, dry summers in extensive high country grasslands. However, the productivity and water use of cocksfoot/lupin pastures is not well understood. The aim of this study was to compare the dry matter (DM) production and water use of cocksfoot/lupin pastures compared with a standard pure lucerne (Medicago sativa) pasture under dryland conditions. The two pasture types were sown in December 2013 at Lincoln University and in this study they were compared during their third year after establishment. During the first and second years after establishment, lucerne produced more DM than cocksfoot/lupin, and the lupin content halved each year such that the cocksfoot dominated the cocksfoot/lupin pasture after 2.5 years. In attempt to increase the legume content of cocksfoot/lupin pasture, a mixture of two cultivars of subterranean clover (Trifolium subterraneum) (Denmark and Narrikup) were oversown in autumn 2017. Six 0.13 ha paddocks of each pasture type were rotationally grazed by young sheep and not irrigated. In the third year after establishment (11 July 2016 to 29 June 2017), lucerne produced 10,617 kg DM/ha compared with 6,433 kg DM/ha for cocksfoot/lupin. The lucerne yield contained 97% lucerne, 0.2% voluntary white clover and 3.2% weed, whereas the cocksfoot/lupin yield contained 86% cocksfoot, 12% lupin, 0.05% sub clover, 1.2% voluntary white clover and 0.8% weed. Sub clover content averaged 2% in the cocksfoot/lupin pasture during the first grazing rotation in the following spring (3 August to 14 September 2017), with 28% of this Denmark and 72% Narrikp. Nitrogen yield in the herbage was 223 kg N/ha for lucerne compared with 87 kg N/ha for cocksfoot/lupin. Pasture growth rates for lucerne and cocksfoot/lupin were 2.06 and 1.21 kg DM/Cd from 11 July to 17 August, then 6.36 and 3.94 kg DM/Cd to 13 December and then 1.75 and 0.88 kg DM/Cd to 29 June. Water use was 503 mm for lucerne and 498 mm for cocksfoot/lupin. Water use efficiency for iii lucerne and cocksfoot/lupin was 6.35 and 3.40 kg DM/ha/mm from 11 July to 17 August, then 23.33 and 16.30 kg DM/ha/mm to 13 December and then 16.46 and 7.50 kg DM/ha/mm to 29 June. Leaf area index, which was quantified pre-grazing in each paddock from 2 March to 22 May, was 2.0 for lucerne, 2.8 for cocksfoot and 1.5 for lupin. Overall the results suggest that the lucerne pasture was able to produce more DM than the cocksfoot/lupin pasture due to a faster pasture growth particularly in spring. This faster growth rate which lead to a higher water use efficiency, deeper access to soil moisture during summer, and possibly a higher LAI which could have resulted in increased light interception. These attributes make lucerne an attractive option for dryland pastures in areas of the South Island high country where soil fertility is unsuitable for lucerne. By comparison, lupin did not persist and Denmark and Narrikup were established from oversowing in the cocksfoot-dominant pasture.