Search

The Florida Water and Climate Alliance (FloridaWCA) is a stakeholder–scientist partnership committed to increasing the relevance of climate science data and tools at time and space scales needed to support decision-making in water resource management, planning, and supply operations in Florida. Since 2010, a group of university researchers, public utility water resource managers and operators, water management district personnel, and local planners have engaged in a sustained collaboration for the development, sharing, and application of cutting-edge research to the practical issues of water management and distribution in the highly urbanized state of Florida. The authors, all members of FloridaWCA, present a case study of the organization’s history, its achievements, and lessons learned at the organizational, scientific/technical, and personal levels. Their goals are to 1) describe how the organizational process has contributed to actionable science based on posing and answering questions of importance; 2) share its scientific impact and technical contributions; 3) demonstrate the value of such a stakeholder–scientist partnership, and 4) identify organizational and structural components that have influenced its effectiveness, including personal reflections. The FloridaWCA, having reached its tenth anniversary, continues to evolve today as a sustained stakeholder–scientist partnership resulting in both guiding researchers of what is applicable in the field (creating an area of research that is useful to society) while also helping the practitioners to push the envelope on the state-of-the practices that can be informed by current research.

In North Florida, increasing nitrogen loads and water quality declines have become a major concern, in part as result of anthropogenic non-point source activities such as agriculture. The main objective of this study was to investigate the effect of irrigation strategies and nitrogen (N) fertility rates on maize biomass, yield and water productivity in sandy soils. The field experiment was conducted 2015–2017 in Live Oak, Florida using a randomized complete block with a split plot design and four replicates. Treatments evaluated five irrigation strategies: (i) GROW, mimicking grower irrigation practices in the region, (ii) SWB, using a soil water balance to schedule irrigation; (iii) SMS, using soil moisture sensors to schedule irrigation; (iv) RED, applying 60% of the GROW treatment; and (v) NON, non-irrigated, and three N fertility rates: (i) low (157 kg N/ha), (ii) medium (247 kg N/ha), and (iii) high (336 kg N/ha). In comparison to GROW, the SWB, SMS and RED irrigation treatments showed no differences in final biomass, N uptake nor grain yield; however, these treatments achieved on average 41, 47, and 36% irrigation reduction, respectively, without impacts on yield during the three maize seasons evaluated. For most of the variables, statistical differences were found between the low and the high N rates, but no differences compared to the medium N rate. A 26% reduction of N fertilizer was achieved using the medium N rate without negative impact on N uptake, biomass nor yield in comparison to the high N fertilization rate. During this experiment, maize N uptake reached a plateau; thus, potential N losses resulted from applications exceeding recommended rates. Furthermore, the implementation of these more efficient irrigation and N fertilizer management strategies reduced irrigation and N fertilizer applications without negative impacts in yield. Thus, these practices may prevent potential N leaching to waterbodies while improving profits.

HighlightsDSSAT simulations of final N uptake, biomass, and yield for a maize-peanut rotational field experiment with three irrigation treatments and three N fertilizer rates had good performance for the irrigated treatments (average nRMSE of 9%) but greater error for the rainfed treatments (average nRMSE of 15%).Experiments and DSSAT simulations demonstrated that N fertilizer and irrigation applications were reduced by 26% and 60%, respectively, when using a 247 kg N ha-1 fertilizer rate and a sensor-based irrigation schedule rather than conventional practices of 336 kg N ha-1 and a calendar-based irrigation method, with no impact on yield.Simulations demonstrated that N leaching during the crop rotation was reduced by 37% when an N fertilizer rate of 247 kg N ha-1 and sensor-based irrigation scheduling were used versus conventional practices.Soil N increased (=15 mg kg-1) when maize and peanut residues decayed and then leached during the fallow season. Cover or cash crops planted immediately after the maize and peanut harvests have potential to take up this N and reduce leaching.Abstract. Nitrogen (N) is an essential element for crop growth and yield; however, excessive N applications not taken up by crops can result in N leaching from the root zone, increasing N loads to waterbodies and leading to a host of environmental problems. The main objective of this study was to simulate water and N balances for a maize-peanut (Zea mays L. and Arachis hypogaea L.) rotational field experiment with three irrigation treatments and three N fertilizer rates. The irrigation treatments consisted of mimicking grower irrigation practices in the region (GROW), using soil moisture sensors to schedule irrigation (SMS), and non-irrigated (NON). The N fertilizer rates were low, medium, and high (157, 247, and 336 kg N ha-1, respectively) for maize with a constant 17 kg ha-1 for all peanut treatments. DSSAT maize genetic coefficients were calibrated using the SMS-high treatment combination under the assumption of no water or N stress. The other eight treatment combinations were used as independent data for model validation of the crop coefficients. All soil hydrologic parameters were specified based on measured values, and default DSSAT peanut genetic coefficients were used with no calibration. For the irrigated treatments, DSSAT models had good performance for N uptake, biomass, and yield (average nRMSE of 8%) and moderate performance for soil water content (average nRMSE of 18%). Soil nitrate RMSE was 21% lower than the standard deviation of the observed data (5.8 vs. 7.2 mg kg-1). For the rainfed treatments, DSSAT had greater error (average nRMSE of 15% for N uptake, biomass, and yield, and average nRMSE of 31% for soil water). Soil nitrate RMSE was 11% greater than the standard deviation of the observed data (8.0 vs. 7.2 mg kg-1), and nRMSE was >30% during the crop rotation. Simulations estimated that N leaching over the crop rotation was reduced by 24% on average when using the 247 kg N ha-1 fertilizer rate compared to 336 kg N ha-1 across the irrigation treatments. Furthermore, N leaching was reduced by 37% when using SMS to schedule irrigation and the 247 kg N ha-1 fertilizer rate for maize and 17 kg N ha-1 for peanut compared to conventional practices (GROW and 336 kg N ha-1 for maize and 17 kg N ha-1 for peanut). Moreover, this management practice reduced N fertilizer use by 26% and irrigation water use by up to 60% without negative impacts on yield. Observed and simulated soil N increased during maize and peanut residue decay, with simulations estimating that this soil N would leach below the root zone during the fallow season. This leaching could potentially be reduced if a cover crop or cash crop were planted between the maize and peanut crops to take up the mineralized N. Keywords: Agricultural best management practices, Bare fallow, BMPs, Maize-peanut rotation, N balance, N fertilization, N leaching, Sandy soils, Sensor-based irrigation scheduling, Water balance.

Effects of future expansion/intensification of irrigated agriculture on groundwater and surface water levels and availability in a semi-arid watershed were evaluated using an integrated hydrologic model (MIKE SHE/MIKE 11) in conjunction with biophysical measurements. Improved water use efficiency, water storage, and energy policy options were evaluated for their ability to sustain the future (2035) increased groundwater withdrawals. Three future withdrawal scenarios (low = 20, medium = 30, high = 50 wells/100 km2/year) based on the historical rate of growth of irrigation wells were formulated. While well drying from falling groundwater levels was limited to drought and consecutive below average rainfall years, under the current (2015) withdrawals, significant increases in frequency and duration (17–97 days/year) of well drying along with 13–26% (19–37 mm) reductions in surface flows were predicted under the future withdrawals. Higher (27–108%) energy demands of existing irrigation pumps due to declining groundwater levels and reduced hydroelectric generation due to decreased surface flows would create a vicious water-food-energy nexus in the future. Crop failure, one of the main causes of farmers’ emotional distress and death in the region, is predicted to exacerbate under the future withdrawal scenarios. Shift to negative net recharge (−63 mm) and early and prolonged drying of wells under the high scenario will reduce the groundwater availability and negatively affect crop production in more than 60% and 90% of cropped areas in the Rabi (November–February) and summer (March–May) seasons, respectively during a drought year. Individual and combined demand (drip irrigation and reduced farm electricity subsidy) and supply (water storage) management options improved groundwater levels and reduced well drying by 55–97 days/year compared to business-as-usual management under the high scenario. The combined management (50% drip conversion, 50% reduction in subsidy, and enhanced water storage) mitigated well drying even during drought and consecutive below average rainfall years under the high scenario. A conservative economic evaluation for management options under the high scenario showed increases in crop production and per farmer annual profits by $987–$1397 during a drought year (average household income = $1520/year). A scale-up of results showed that diverting 50% state power subsidy ($6 billion for 3–6 years) can almost entirely fund the conversion to drip irrigation ($4.2 billion) and water storage structures ($2.9 billion) and help meet the water supply demand of a 50% increase in irrigated area under the high scenario. Converting flood to drip irrigation in 50% of irrigated area under the high scenario can reduce the electric energy consumption (7 × 106Mwh/year) and carbon footprint (6000 Mt/year) of groundwater irrigation by 24% in the state. Management options considered can potentially create a sustainable water-food-energy nexus in the larger semi-arid hard rock region. Reducing the power subsidy will require a strong political will since it has been used as a tool to win the elections in India. Considering future agricultural intensification, timely interventions are needed to ensure the livelihood and well-being of millions of small- and medium-scale farmers that rely on low storage, hard rock aquifers in the semi-arid regions of the world.

In this paper we present a hydro-chemical model that can be used to generate plausible karstic conduit networks that honor what is known about geology, hydrology and topography of a karst system. To make the model applicable to a range of natural karst systems, we introduce a flexible and physically realistic flow boundary condition along the land surface. Moreover, whereas comparable existing speleogenesis models use an explicit reactive-transport scheme, we propose an implicit reactive-transport scheme to permit a coarser spatial discretization of the conduit cells. An application to a real karst system illustrates that the model can generate a realistic karstic network that reproduces observed hydrologic behavior in terms of current spring flow rates, regional hydraulic head field as well as average groundwater residence times. Our model provides a useful tool to generate ensembles of possible karstic conduit networks that may be used within a stochastic framework to analyze flow and transport prediction uncertainty associated with a lack of knowledge about network geometry.

This study integrates push-pull tracer tests (PPTT) with microbial characterization of extracted water via quantitative polymerase chain reaction (qPCR) and reverse transcriptase qPCR (RT-qPCR) of selected functional N transformation genes to quantify nitrate reduction mechanisms and rates in sites with different redox potential in a karst aquifer. PPTT treatments with nitrate (AN) and nitrate-fumarate (ANC) were executed in two wells representing anoxic and oxic geochemical end-members. Oxic aquifer zero-order nitrate loss rates (mmol L−1 h−1) were similar for AN and ANC treatment, ranging from 0.03 ± 0.01 to 0.05 ± 0.01. Anoxic aquifer zero-order nitrate loss rates ranged from 0.03 ± 0.02 (AN) to 0.13 ± 0.02 (ANC). Microbial characterization indicates mechanisms influencing these rates were dissimilatory nitrate reduction to ammonium (DNRA) at the anoxic site with AN treatment, assimilatory reduction of nitrate to ammonium (ANRA) with ANC treatment in the water column at both sites, and additional documented nitrate reduction that occurred in unsampled biofilms. With carbon treatment, total numbers of microbes (16S rRNA genes) significantly increased (fourteenfold to thirtyfold), supporting stimulated growth with resulting ANRA. Decreased DNRA gene concentrations (nrfA DNA) and increased DNRA activity ratio (nrfA-cDNA/DNA) supported the assertion that DNRA occurred in the anoxic zone with AN and ANC treatment. Furthermore, decreased DNRA gene copy numbers at the anoxic site with ANC treatment suggests that DNRA microbes in the anoxic site are chemolithoautotrophic. Increased RT-qPCR denitrification gene expression (nirK and nirS) was not observed in water samples, supporting that any observed NO3-N loss due to denitrification may be occurring in unsampled microbial biofilms.

Nitrate leaching from agricultural fields is a significant contributor of groundwater pollution globally, threatening drinking water resources and downstream ecosystems. Quantifying nitrate leaching driven by variable climate, soils, and management practices is challenging, but it is critical for developing sustainable agricultural production systems. While irrigation and fertilizer “best management practices” (BMPs) have been widely implemented to reduce agricultural nitrate leaching, their ability to meet environmental protection goals remains uncertain. In this study, we used the Soil and Water Assessment Tool (SWAT) to simulate crop yields and nitrate leaching for corn-peanut rotations under a variety of nutrient and irrigation management practices in the Suwannee River Basin (Florida), where groundwater feeds springs that are protected by a federally mandated nutrient criteria of 0.35 mg/L Nitrate-Nitrogen (NO3-N). Data from a field experiment of nine irrigation and nitrogen (N) management treatments were used to calibrate SWAT, with good to excellent results (Nash Sutcliffe Efficiencies from 0.72 to 0.97 for soil moisture, 0.85–0.96 for crop yield, 0.48–0.96 for crop N uptake, and 0.15–0.82 for soil nitrate). The calibrated model was then used to quantify differences in crop yields, irrigation applied and nitrate leaching among practices over a range of historical weather. Soil moisture sensor-based irrigation with 246 kg N/ha for corn and 0 kg N/ha for peanut had no statistical difference in yields compared to common practices in the region (calendar-based irrigation, fertilization of 336 kg N/ha corn and 17 kg N/ha peanut), while reducing N leaching by 40% and irrigation applied by 45% (reductions of ~70 kg N/ha/yr and ~300 mm/year, respectively). Planting a rye cover crop reduced leaching by an additional ~50 N/ha/yr for all treatments. These results show the potential for widespread adoption of nutrient and water conservation practices to achieve the reductions in NO3-N load needed to meet environmental and regulatory goals without impacting crop yields.

General circulation models (GCMs) have been widely used to simulate current and future climate at the global scale. However, the development of frameworks to apply GCMs to assess potential climate change impacts on regional hydrologic systems, ability to meet future water demand, and compliance with water resource regulations is more recent. In this study eight GCMs were bias-corrected and downscaled using the bias correction and stochastic analog (BCSA) downscaling method and then used, together with three ET0 methods and eight different water use scenarios, to drive an integrated hydrologic model previously developed for the Tampa Bay region in western central Florida. Variance-based sensitivity analysis showed that changes in projected streamflow were very sensitive to GCM selection, but relatively insensitive to ET0 method or water use scenario. Changes in projections of groundwater level were sensitive to both GCM and water use scenario, but relatively insensitive to ET0 method. Five of eight GCMs projected a decrease in streamflow and groundwater availability in the future regardless of water use scenario or ET method. For the business as usual water use scenario all eight GCMs indicated that, even with active water conservation programs, increases in public water demand projected for 2045 could not be met from ground and surface water supplies while achieving current groundwater level and surface water flow regulations. With adoption of 40 % wastewater reuse for public supply and active conservation four of the eight GCMs indicate that 2045 public water demand could be met while achieving current environmental regulations; however, drier climates would require a switch from groundwater to surface water use. These results indicate a high probability of a reduction in future freshwater supply in the Tampa Bay region if environmental regulations intended to protect current aquatic ecosystems do not adapt to the changing climate. Broad interpretation of the results of this study may be limited by the fact that all future water use scenarios assumed that increases in water demand would be the result of intensification of water use on existing agricultural, industrial, and urban lands. Future work should evaluate the impacts of a range of potential land use change scenarios, with associated water use change projections, over a larger number of GCMs.

Projecting water deficit under various possible future climate scenarios depends on the choice of general circulation model (GCM), reference evapotranspiration (ET0) estimation method, and Representative Concentration Pathway (RCP) trajectory. The relative contribution of each of these factors must be evaluated in order to choose an appropriate ensemble of future scenarios for water resources planning. In this study variance-based global sensitivity analysis and Monte Carlo filtering were used to evaluate the relative sensitivity of projected changes in precipitation (P), ET0, and water deficit (defined here as P–ET0) to choice of GCM, ET0 estimation method, and RCP trajectory over the continental United States (US) for two distinct future periods: 2030–2060 (future period 1) and 2070–2100 (future period 2). A total of 9 GCMs, 10 ET0 methods, and 3 RCP trajectories were used to quantify the range of future projections and estimate the relative sensitivity of future projections to each of these factors. In general, for all regions of the continental US, changes in future precipitation are most sensitive to the choice of GCM, while changes in future ET0 are most sensitive to the choice of ET0 estimation method. For changes in future water deficit, the choice of GCM is the most influential factor in the cool season (December–March), and the choice of ET0 estimation method is most important in the warm season (May–October) for all regions except the Southeast US, where GCMs and ET0 have approximately equal influence throughout most of the year. Although the choice of RCP trajectory is generally less important than the choice of GCM or ET0 method, the impact of RCP trajectory increases in future period 2 over future period 1 for all factors. Monte Carlo filtering results indicate that particular GCMs and ET0 methods drive the projection of wetter or drier future conditions much more than RCP trajectory; however, the set of GCMs and ET0 methods that produce wetter or drier projections varies substantially by region. Results of this study indicate that, in addition to using an ensemble of GCMs and several RCP trajectories, a range of regionally relevant ET0 estimation methods should be used to develop a robust range of future conditions for water resources planning under climate change.

There are a number of statistical techniques that downscale coarse climate information from global circulation models (GCM). However, many of them do not reproduce the small-scale spatial variability of precipitation exhibited by the observed meteorological data which can be an important factor for predicting hydrologic response to climatic forcing. In this study a new downscaling technique (bias-correction and stochastic analog method, BCSA) was developed to produce stochastic realizations of bias-corrected daily GCM precipitation fields that preserve the spatial autocorrelation structure of observed daily precipitation sequences. This approach was designed to reproduce observed spatial and temporal variability as well as mean climatology. We used the BCSA method to downscale 4 GCM precipitation predictions from 1961 to 1999 over the state of Florida and compared the skill of the method to the results obtained with the commonly used bias-correction and spatial disaggregation (BCSD) approach, bias-correction and constructed analog (BCCA) method, and a modified version of BCSD which reverses the order of spatial disaggregation and bias-correction (SDBC). Spatial and temporal statistics, transition probabilities, wet/dry spell lengths, spatial correlation indices, and variograms for wet (June through September) and dry (October through May) seasons were calculated for each method. Results showed that (1) BCCA underestimated mean climatology of daily precipitation while the BCSD, SDBC and BCSA methods accurately reproduced it, (2) the BCSD and BCCA methods underestimated temporal variability because of the interpolation and regression schemes used for downscaling and thus, did not reproduce daily precipitation standard deviations, transition probabilities or wet/dry spell lengths as well as the SDBC and BCSA methods, and (3) the BCSD, BCCA and SDBC methods underestimated spatial variability in precipitation resulting in under-prediction of spatial variance and over-prediction of spatial correlation, whereas the new stochastic technique (BCSA) accurately reproduces observed spatial statistics for both the wet and dry seasons. This study underscores the need to carefully select a downscaling method that reproduces all precipitation characteristics important for the hydrologic system under consideration if local hydrologic impacts of climate variability and change are going to be accurately predicted. For low-relief, rainfall-dominated watersheds where reproducing small-scale spatiotemporal precipitation variability is important, the BCSA method is recommended for use over the BCSD, BCCA, or SDBC methods.

General circulation models (GCMs) have been widely used to simulate current and future climate at the global scale. However, the development of frameworks to apply GCMs to assess potential climate change impacts on regional hydrologic systems and compliance with water resource regulations is more recent. It is important to predict potential impacts of future climate change on streamflows and groundwater levels to reduce risks and increase resilience in water resources management and planning. This study evaluated future streamflows and groundwater levels in the Tampa Bay region in west-central Florida using an ensemble of different GCMs, reference evapotranspiration (ET0) methods, and water use scenarios to drive an integrated hydrologic model (IHM). Eight GCMs were bias-corrected and downscaled using the Bias Correction and Stochastic Analog (BCSA) downscaling method and then used, together with three ET0 methods, to drive the IHM for eight different human water use scenarios. Results showed that changes in projected streamflow were most sensitive to GCM selection, however, projections of groundwater level change were sensitive to both GCM and water use scenario. Projected changes in streamflow and groundwater level were relatively insensitive to the ET0 methods evaluated in this study. Six of eight GCMs projected a decrease in streamflow and groundwater level in the future regardless of water use scenario or ET method. These results indicate a high probability of a reduction in future water supply in the Tampa Bay region if environmental regulations intended to protect current aquatic ecosystems do not adapt to the changing climate.

Simultaneous effects of future climate and irrigation intensification on surface and groundwater systems are not well understood. Efforts are needed to understand the future groundwater availability and associated surface flows under business-as-usual management to formulate policy changes to improve water sustainability. We combine measurements with integrated modeling (MIKE SHE/MIKE11) to evaluate the effects of future climate (2040–2069), with and without irrigation expansion, on water levels and flows in an agricultural watershed in low-storage crystalline aquifer region of south India. Demand and supply management changes, including improved efficiency of irrigation water as well as energy uses, were evaluated. Increased future rainfall (7–43%, from 5 Global Climate Models) with no further expansion of irrigation wells increased the groundwater recharge (10–55%); however, most of the recharge moved out of watershed as increased baseflow (17–154%) with a small increase in net recharge (+0.2 mm/year). When increased rainfall was considered with projected increase in irrigation withdrawals, both hydrologic extremes of well drying and flooding were predicted. A 100-year flow event was predicted to be a 5-year event in the future. If irrigation expansion follows the historical trends, earlier and more frequent well drying, a source of farmers' distress in India, was predicted to worsen in the future despite the recharge gains from increased rainfall. Storage and use of excess flows, improved irrigation efficiency with flood to drip conversion in 25% of irrigated area, and reduced energy subsidy (free electricity for 3.5 h compared to 7 h/day; $1 billion savings) provided sufficient water savings to support future expansion in irrigated areas while mitigating well drying as well as flooding. Reductions in energy subsidy to fund the implementation of economically desirable (high benefit-cost ratio) demand (drip irrigation) and supply (water capture and storage) management was recommended to achieve a sustainable food-water-energy nexus in semi-arid regions.

Crop models are essential tools in climate change impact assessments, but they often lack comprehensive field testing. In this study, we tested the SUBSTOR-potato model with 87 field experiments, including 204 treatments from 19 countries. The field experiments varied in potato species and cultivars, N fertilizer application, water supply, sowing dates, soil types, temperature environments, and atmospheric CO2 concentrations, and included open top chamber and Free-Air-CO2-Enrichment (FACE) experiments. Tuber yields were generally well simulated with the SUBSTOR-potato model across a wide range of current growing conditions and for diverse potato species and cultivars, including Solanum tuberosum, Solanum andigenum, Solanum juzepczukii species, as well as modern, traditional, early, medium, and late maturity-type cultivars, with a relative RMSE of 37.2% for tuber dry weight and 21.4% for tuber fresh weight. Cultivars 'Desiree' and 'Atlantic' were grown in experiments across the globe and well simulated using consistent cultivar parameters. However, the model underestimated the impact of elevated atmospheric CO2 concentrations and poorly simulated high temperature effects on crop growth. Other simulated crop variables, including leaf area, stem weight, crop N, and soil water, differed frequently from measurements; some of these variables had significant large measurement errors. The SUBSTOR-potato model was shown to be suitable to simulate tuber growth and yields over a wide range of current growing conditions and crop management practices across many geographic regions. However, before the model can be used effectively in climate change impact assessments, it requires improved model routines to capture the impacts of elevated atmospheric CO2 and high temperatures on crop growth.

This study compared two types of approaches to downscale seasonal precipitation (P) and 2-m air temperature (T2M) forecasts from the North American Multimodel Ensemble (NMME) over the states of Alabama, Georgia, and Florida in the southeastern United States (SEUS). Each NMME model forecast was evaluated. Two multimodel ensemble (MME) schemes were tested by assigning equal weight to all forecast members (SuperEns) or by assigning equal weights to each model’s ensemble mean (MeanEns). One type of downscaling approach used was a model output statistics (MOS) method, which was based on direct spatial disaggregation and bias correction of the NMME P and T2M forecasts using the quantile mapping technique [spatial disaggregation with bias correction (SDBC)]. The other type of approach used was a perfect prognosis (PP) approach using nonparametric locally weighted polynomial regression (LWPR) models, which used the NMME forecasts of Niño-3.4 sea surface temperatures (SSTs) to predict local-scale P and T2M. Both SDBC and LWPR downscaled P showed skill in winter but no skill or limited skill in summer at all lead times for all NMME models. The SDBC downscaled T2M were skillful only for the Climate Forecast System, version 2 (CFSv2), model even at far lead times, whereas the LWPR downscaled T2M showed limited skill or no skill for all NMME models. In many cases, the LWPR method showed significantly higher skill than the SDBC. After bias correction, the SuperEns mostly showed higher skill than the MeanEns and most of the single models, but its skill did not outperform the best single model.

Summary This study demonstrated the first application of a GSA technique to a transient ISSHM–LSM application developed for a large-scale river basin. The Morris method was used to identify the spatially and temporally variable sensitivity amongst a large number of model parameters to provide insights on hydrologic processes dominating behavior in the basin and to identify a small subset of parameters that should be evaluated in subsequent, more computationally intensive quantitative GSA and parameter estimation techniques. Results showed that in the upper region of the basin, evapotranspiration (ET), total streamflow and peak streamflow were less sensitive to surficial aquifer system characteristics, but highly sensitive to the hydraulic conductivity of the confining unit separating the surficial aquifer and the regional aquifer system and leaf area index of near stream vegetation. In the lower region of the basin, hydraulic conductivity of the regional aquifer system was found to have a significant effect on ET, total stream flow, and groundwater contributions to streamflow while surface–groundwater dynamics during storm events was most sensitive to storage properties of the regional aquifer system. Peak streamflow in the lower basin was most sensitive to the hydraulic conductivity of the confining unit in the upper basin, and the Manning’s coefficient of upper basin streams, indicating that all peak storm flows originate in the upper basin. Throughout the basin ET was sensitive to soil/geologic properties and vegetation properties, with unsaturated zone processes and relevant parameters gaining importance in moisture limited conditions existing in the lower regions of the basin.

Reference evapotranspiration (ETo) is an important hydroclimatic variable for water planning and management. This research explored the potential of using the Climate Forecast System, version 2 (CFSv2), for seasonal predictions of ETo over the states of Alabama, Georgia, and Florida. The 12-km ETo forecasts were produced by downscaling coarse-scale ETo forecasts from the CFSv2 retrospective forecast archive and by downscaling CFSv2 maximum temperature (Tmax), minimum temperature (Tmin), mean temperature (Tmean), solar radiation (Rs), and wind speed (Wind) individually and calculating ETo using those downscaled variables. All the ETo forecasts were calculated using the Penman–Monteith equation. Sensitivity coefficients were evaluated to quantify how and how much does each of the variables influence ETo. Two statistical downscaling methods were tested: 1) spatial disaggregation (SD) and 2) spatial disaggregation with quantile mapping bias correction (SDBC). The downscaled ETo from the coarse-scale ETo showed similar skill to those by first downscaling individual variables and then calculating ETo. The sensitivity coefficients showed Tmax and Rs had the greatest influence on ETo, followed by Tmin and Tmean, and Wind. The downscaled Tmax showed highest predictability, followed by Tmean, Tmin, Rs, and Wind. SDBC had slightly better performance than SD for both probabilistic and deterministic forecasts. The skill was locally and seasonally dependent. The CFSv2-based ETo forecasts showed higher predictability in cold seasons than in warm seasons. The CFSv2 model could better predict ETo in cold seasons during El Niño–Southern Oscillation (ENSO) events only when the forecast initial condition was in either the El Niño or La Niña phase of ENSO.

Subsurface heterogeneity is one of the largest sources of uncertainty associated with saturated hydraulic conductivity. Recent work has demonstrated that uncertainty in hydraulic conductivity can impart significant uncertainty in runoff generation processes and surface-water flow. Here, the role of site characterization in reducing hydrograph prediction bias and uncertainty is demonstrated. A fully integrated hydrologic model is used to conduct two sets of stochastic, transient simulation experiments comprising different overland flow mechanisms: Dunne and Hortonian. Conditioning hydraulic conductivity fields using values drawn from a simulated synthetic control case are shown to reduce both mean bias and variance in an ensemble of conditional hydrograph predictions when compared with the control case. The ensemble simulations show a greater reduction in uncertainty in the hydrographs for Hortonian flow. The conditional simulations predict surface ponding and surface pressure distributions with reduced mean error and reduced root mean square error compared with unconditional simulations. Uncertainty reduction in Hortonian and Dunne flow cases demonstrates different temporal signals, with more substantial reduction achieved for Hortonian flow.

This study applied three statistical downscaling methods: (1) bias correction and spatial disaggregation at daily time scale (BCSD_daily); (2) a modified version of BCSD which reverses the order of spatial disaggregation and bias correction (SDBC), and (3) the bias correction and stochastic analog method (BCSA) to downscale general circulation model daily precipitation outputs to the subbasin scale for west-central Florida. Each downscaled climate input dataset was then used in an integrated hydrologic model to examine differences in ability to simulate retrospective streamflow characteristics. Results showed the BCSD_daily method consistently underestimated mean streamflow because the highly spatially correlated small precipitation events produced by this method resulted in overestimation of evapotranspiration. Highly spatially correlated large precipitation events produced by the SDBC method resulted in overestimation of the standard deviation of wet season daily streamflow and the magnitude/frequency of high streamflow events. BCSA showed better performance than the other methods in reproducing spatiotemporal statistics of daily precipitation and streamflow. This study demonstrated differences in statistical downscaling techniques propagate into significant differences in streamflow predictions, and underscores the need to carefully select a downscaling method that reproduces precipitation characteristics important for the hydrologic system under consideration.

Groundwater flow in karst includes exchange of water between large fractures, conduits, and the surrounding porous matrix, which impacts both water quality and quantity. Electrical resistivity tomography combined with end-member mixing analysis (EMMA) and numerical flow and transport modeling was used to study mixing of karst conduit and matrix waters to understand spatial and temporal patterns of mixing during high flow and base flow conditions. To our knowledge, this is the first time EMMA and synthetic geophysical simulations have been combined. Here we interpret an 8 week time-lapse electrical resistivity data set to assess groundwater-surface mixing. We simulate flow between the karst conduits and the porous matrix to determine fractions of water recharged to conduits that has mixed with groundwater stored in the pore space of the matrix using a flow and transport model in a synthetic time-lapse resistivity inversion. Comparing the field and synthetic inversions, our results enable us to estimate exchange dynamics, spatial mixing, and flow conditions. Results showed that mixing occurred at a volumetric flux of 56 m3/d with a dispersivity around 1.69 m during the geophysical experiment. For these conditions, it was determined that conduit water composition ranged from 75% groundwater during base flow conditions to less than 50% groundwater in high flow conditions. Though subject to some uncertainties, the time-lapse inversion process provides a means to predict changing hydrologic conditions, leading to mixing of surface water and ground water and thus changes to water quantity and quality, as well as potential for water-rock reactions, in a semiconfined, sink-rise system.

Summary This study investigated the limitations associated with using dynamically-downscaled, bias-corrected reanalysis data (i.e. regional reanalysis data) to predict hydrologic behavior of low-relief rainfall driven systems using an integrated surface/groundwater model. Four different sets of global reanalysis data (NCEP/NCAR-R1, NCEP-DOE-R2, ERA40, and 20CR) that were previously downscaled using two RCMs (MM5 and RSM) were obtained, bias-corrected on a daily basis using the CDF-mapping approach, and used to drive an integrated hydrologic model (INTB) that was previously calibrated and verified for the Tampa Bay region. All raw dynamically-downscaled reanalysis datasets accurately estimated the annual cycle of daily maximum and minimum temperature, except the NCEP/NCAR R1+MM5 data which consistently underestimated daily maximum temperature. All raw regional reanalysis precipitation data significantly overestimated precipitation, particularly for the dry season. Bias-correction using the CDF-mapping approach effectively removed biases in the temporal mean and standard deviation of both the daily precipitation and temperature predictions. Biases in the mean monthly and mean annual precipitation totals were removed by CDF-mapping on a daily basis, but the standard deviation of the monthly and annual precipitation totals were not accurately reproduced. Furthermore inaccuracies in actual daily precipitation time series aggregated into monthly and annual rainfall total time series that showed significant and temporally persistent errors. Precipitation timing errors produced by regional reanalysis data were propagated and enhanced by non-linear streamflow generation, groundwater flow and storage processes in the hydrologic model and produced significant errors in both actual and mean daily, monthly and annual streamflow and groundwater level predictions. These results show that improvement in large-scale reanalysis products and regional climate models may be required before dynamically downscaled bias-corrected reanalysis data can be used as a surrogate for observational data in hydrologic model applications for low-relief, rainfall driven systems.

Physics-based distributed models for simulating flow in karst systems are generally based on the discrete–continuum approach in which the flow in the three-dimensional fractured limestone matrix continuum is coupled with the flow in discrete one-dimensional conduits. In this study we present a newly designed discrete–continuum model for simulating flow in karst systems. We use a flexible spatial discretization such that complicated conduit networks can be incorporated. Turbulent conduit flow and turbulent surface flow are described by the diffusion wave equation whereas laminar variably saturated flow in the matrix is described by the Richards equation. Transients between free-surface and pressurized conduit flow are handled by changing the capacity term of the conduit flow equation. This new approach has the advantage that the transients in mixed conduit flow regimes can be handled without the Preissmann slot approach. Conduit–matrix coupling is based on the Peaceman’s well-index such that simulated exchange fluxes across the conduit–matrix interface are less sensitive to the spatial discretization. Coupling with the surface flow domain is based on numerical techniques commonly used in surface–subsurface models and storm water drainage models. Robust algorithms are used to simulate the non-linear flow processes in a coupled fashion. The model is verified and illustrated with simulation examples.

A Lagrangian particle tracking scheme has been extended to simulate advective transport through coupled surface and subsurface flows. This extended scheme assumes a two-dimensional flow field for the overland domain and a three-dimensional flow field for the subsurface domain. Moreover it is assumed that the flow fields are simulated by a cell centered finite difference method. Pathlines in both the subsurface and the overland domain are simulated by classical particle tracking methods. Exchange of particles between the two domains is simulated by newly-developed algorithms presented in this study. Different algorithms are used depending on the direction of the exchange across the interface between the two domains. In the subsurface domain knowledge about a particle’s pathline is enough to detect a transfer to the surface domain and the solution is straightforward. However, in the two-dimensional overland domain pathlines are parallel to the land surface. Therefore the velocity field in the overland domain alone cannot be used to detect a transfer to the subsurface. We propose a relatively simple algorithm to estimate the probability of transfer to the subsurface domain. It is shown that this algorithm can also be used to handle the transfer from the overland domain to the atmosphere domain. The algorithm to estimate the transfer probabilities is based on the mass balance of water on a streamtube aligned with the particle’s pathline. This newly developed technique ensures that transit time distributions can be simulated accurately. These new relationships are implemented in an existing particle tracking code and are verified using analytical solutions for transit times.

The goal of this study was to evaluate the ability of dynamically downscaled reanalysis data to reproduce local-scale spatiotemporal precipitation and temperature data needed to accurately predict streamflow in the Tampa Bay region of west central Florida. In particular, the Florida State University Center for Ocean-Atmospheric Prediction Studies CLARReS10 data (NCEP DOE 2 reanalysis data (R2) downscaled to 10-km over the Southeast USA using the Regional Spectral Model (RSM) were evaluated against locally available observed precipitation and temperature data and then used to drive an integrated hydrologic model that was previously calibrated for the Tampa Bay region. Resulting streamflow simulations were evaluated against observed data and previously calibrated model results. Results showed that the raw downscaled reanalysis predictions accurately reproduced the seasonal trends of mean daily minimum temperature, maximum temperature and precipitation, but generally overestimated the monthly mean and standard deviation of daily precipitation. Biases in the temporal mean and standard deviation of daily precipitation and temperature predictions were effectively removed using a CDF-mapping approach; however, errors in monthly precipitation totals remained after bias correction. Monthly streamflow simulation error statistics indicated that the accuracy of the streamflow produced by the bias-corrected downscaled reanalysis data was satisfactory (i.e., sufficient for seasonal to decadal planning), but that the accuracy of the streamflow produced by the raw downscaled reanalysis data was unsatisfactory for water resource planning purposes. The findings of this study thus indicate that further improvement in large-scale reanalysis data and regional climate models is needed before dynamically downscaled reanalysis data can be used directly (i.e. without bias correction with local data) to drive hydrologic models. However, bias-corrected dynamically downscaled data show promise for extending local historic climate observation records for hydrologic simulations. Furthermore, results of this study indicate that similarly bias-corrected dynamically downscaled retrospective and future GCM projections should be suitable for assessing potential hydrologic impacts of future climate change in the Tampa Bay region.

Projecting water availability under various possible future climate scenarios depends on the choice of Global Climate Model (GCM), evapotranspiration (ET) estimation method and Representative Concentration Pathway (RCP) trajectory. The relative contribution of each of these factors must be evaluated in order to choose an appropriate ensemble of future scenarios for water resources planning. In this study variance-based global sensitivity analysis and Monte Carlo filtering were used to evaluate the relative sensitivity of projected changes in precipitation (P), ET and water availability (defined here as P–ET) to choice of GCM, ET estimation method and RCP trajectory over the continental United States (US) for two distinct future periods: 2030–2060 (future period 1) and 2070–2100 (future period 2). A total of 9 GCMs, 10 ET methods and 3 RCP trajectories were used to quantify the range of future projections and estimate the relative sensitivity of future projections to each of these factors. In general, for all regions of the US, changes in future precipitation are most sensitive to the choice of GCM, while changes in future ET are most sensitive to the choice of ET estimation method. For changes in future water availability, the choice of GCM is the most influential factor in the cool season (December–March) and the choice of ET estimation method is most important in the warm season (May–October) for all regions except the South East US where GCM and ET have approximately equal influence throughout most of the year. Although the choice of RCP trajectory is generally less important than the choice of GCM or ET method, the impact of RCP trajectory increases in future period 2 over future period 1 for all factors. Monte Carlo filtering results indicate that particular GCMs and ET methods drive the projection of wetter or drier future conditions much more than RCP trajectory; however the set of GCMs and ET methods that produce wetter or drier projections varies substantially by region. Results of this study indicate that, in addition to using an ensemble of GCMs and several RCP trajectories, a range of regionally-relevant ET estimation methods should be used to develop a robust range of future conditions for water resource planning under climate change.

In this paper, L-band microwave observations were assimilated using the ensemble Kalman filter to improve root-zone soil moisture (RZSM) estimates from a coupled soil vegetation atmosphere transfer (SVAT)-vegetation model linked to a forward microwave model. Simultaneous state-parameter updates were performed by assimilating both synthetic and field observations during a growing season of sweet corn every three days, matching the temporal coverage of observations from the Soil Moisture and Ocean Salinity and Soil Moisture Active Passive missions. The sensitivities of parameters to the states were investigated using the information-theoretic measure of conditional entropy. Among the soil parameters, the pore-size index (λ) was the most sensitive to brightness temperatures (TB) during the early and midgrowth stages, while porosity (φ) was the most sensitive to TB during the reproductive stage. In the microwave model, the soil roughness parameters, root mean square (RMS) height (r), and correlation length (l) were the most sensitive during the early and mid stages, while the vegetation regression parameter (b) was the most sensitive during the reproductive stage. In the synthetic experiment, assimilation of TB provided RMS error reductions in RZSM estimates of 70% compared to open loop estimates. Minimal variations in performance were observed across different stages of the season during the synthetic experiment. However, when field observations of TB were assimilated, significant differences in RZSM estimates were observed during different growth stages. Maximum RMS difference (RMSD) reductions in RZSM estimates of 33.3% were observed compared to open loop estimates during the early stages, while improvements of 4.8% and 16.7% were observed in the mid- and reproductive stages, respectively. Further analyses of assimilation with field observations also suggest some improvements in the SVAT model are needed for moisture transport immediately following the precipitation/irrigation events. In the microwave model, the linear vegetation formulation for estimating canopy opacity, parameterized by b, was inadequate in capturing the complexities in TB during stages of high vegetative and reproductive growth rates.

Research based crop-specific best management practices (BMPs) must be developed for sweet corn ( Zea mays L. var. saccharata ) production to reduce the amount of nitrogen (N) leaching. The objective of this study was to identify irrigation and nitrogen BMPs for sweet corn production on sandy soils in Florida using the calibrated CERES-Maize model of the Decision Support System for Agrotechnology Transfer (DSSAT). A total of 24 irrigation schedules, 21 N fertilizer levels, 30 N application splits, and 20 N application rates per split were systematically evaluated in single factor simulations. Then, a set of 324 management scenarios composed of 6 irrigation timing/amount and 54 N fertilizer application strategies selected in early single factor explorations, was explored in a multifactor analysis. Irrigation frequency had a strong influence on sweet corn yield. If irrigation events were triggered when maximum allowable depletion (MAD) of soil water content was greater than 60%, corn growth suffered water stress and the simulated yield was reduced. The increase in yield approached zero above 168 kg N ha −1 . Splitting N fertilizer applications did not influence yield if there was an N application during the small-leaf stage or large-leaf stage; however, the lowest amount of N leaching occurred when no N was applied during the small-leaf stage. Simulated yield increased when application rates decreased from 100 to 70 kg N ha −1 per fertigation event, but changed only slightly at application rates less than 70 kg N ha −1 per fertigation. Smaller application rates per fertigation decreased N leaching substantially, especially for rates less than 70 kg N ha −1 . Six potential BMPs were selected from the 324 management scenarios as optimizing yield while minimizing N leaching. These BMPs were composed of two irrigation schedules (depths of 5.0 and 7.5 mm with MAD values of 20% and 30%), two N levels (196 and 224 kg N ha −1 ), two N split plans (0-1/4-3/4 and 0-1/3-2/3 of total N applied in the small-leaf, large-leaf, and ear development stages, respectively), and two N application rates per fertigation (30 and 40 kg N ha −1 ). It should be recognized that these results are recommendations based on modeling assumptions and should be tested in actual field production for their practical and economic validity.

Pastures in beef cattle ranches are identified as one of the major land uses that contribute phosphorus (P) and nitrogen (N) loadings to Lake Okeechobee (LO), Florida. A study was conducted to quantify flow, P, and N contributions from groundwater to a 170 m ditch section within a beef cattle ranch located in the LO basin from June 2005 to May 2009. Groundwater depths and total N (TN) and total P (TP) concentrations measured in five wells, installed on each side (bank) of the ditch section, were combined with the ditch stage and N and P concentrations to quantify groundwater flow and nutrient loads. The Dupuit equation was used to estimate the groundwater flow to and from the ditch. The average net groundwater flow to the ditch section, estimated from a range (low and high) of saturated hydraulic conductivity (Ksat) values, ranged from 765 to 1,729 m3 per year. Annual groundwater flow contributions were positive (to the ditch) for all four years, while the groundwater nutrient contributions were positive for three years and negative (from the ditch) for one year (drought year). Average net annual TP loads to the ditch section ranged from 0.1 (low) to 0.2 (high) kg, while TN loads ranged from 2.1 (low) to 4.5 (high) kg. Extrapolating the results from this study to the LO basin indicated that groundwater was a major contributor of TP loads to the lake.

This research quantitatively evaluated the ability of the fifth-generation Pennsylvania State University–National Center for Atmospheric Research Mesoscale Model (MM5) to reproduce observed spatiotemporal variability of precipitation in the Tampa Bay region over the 1986–2008 period. Raw MM5 model results were positively biased; therefore, the raw model precipitation outputs were bias corrected at 53 long-term precipitation stations in the region using the cumulative distribution function (CDF) mapping approach. CDF mapping effectively removed the bias in the mean daily, monthly, and annual precipitation totals and improved the RMSE of these rainfall totals. Observed daily precipitation transition probabilities were also well predicted by the bias-corrected MM5 results. Nevertheless, significant error remained in predicting specific daily, monthly, and annual total time series. After bias correction, MM5 successfully reproduced seasonal geostatistical precipitation patterns, with higher spatial variance of daily precipitation in the wet season and lower spatial variance of daily precipitation in the dry season. Bias-corrected daily precipitation fields were kriged over the study area to produce spatiotemporally distributed precipitation fields over the dense grids needed to drive hydrologic models in the Tampa Bay region. Cross validation at the 53 long-term precipitation gauges showed that kriging reproduced observed rainfall with average RMSEs lower than the RMSEs of individually bias-corrected point predictions. Results indicate that although significant error remains in predicting actual daily precipitation at rain gauges, kriging the bias-corrected MM5 predictions over a hydrologic model grid produces distributed precipitation fields with sufficient realism in the daily, seasonal, and interannual patterns to be useful for multidecadal water resource planning in the Tampa Bay region.

Ditch fencing and culvert cattle crossing Best Management Practice (BMP) was evaluated in this study with regard to phosphorus (P) and nitrogen (N) load reductions and economic feasibility in the Lake Okeechobee (LO) basin. The BMP was implemented at a 170 m section of a drainage ditch within a ranch in the LO basin and flow and concentration (N and P) data at the upstream and downstream of the ditch were collected for one pre-BMP (June–October, 2005) and three post-BMP (June–October, 2006–2008) periods. During the pre-BMP period, downstream total P (TP) load was 20% (67.0 kg) higher than the upstream, indicating the cattle crossing to be a source of P. Downstream loads of TP in 2006 and 2008 (post-BMP periods) became 26% (14.7 kg) and 11% (85.9 kg) lower than the upstream loads, respectively indicating that the BMP reduced the P loads. The site was a sink for N for all periods except the 2007. Unusual dry conditions during 2007 resulted in the addition of P and N at the BMP site, probably due to the release of P and N from soil and plants. Average of three post-BMP period load showed a 10% reduction of TP loads at the downstream (251.8 kg) compared to the upstream (281.0 kg) location. To consider potential P contributions from the soil and plant, two scenarios, conservative and liberal, were considered to estimate P load reductions due to the BMP. For the conservative scenario, P contribution from soil and plant was considered, while for liberal it was not. Reductions in P loads for conservative and liberal scenarios were 0.35 and 0.44 kg/day, respectively. Phosphorus removal cost for the conservative scenario was $12.61/kg of P, which is considerably less than the cost of other P reduction strategies in the basin. Overall, results show that the BMP can reduce P concentration and loads from ranches without causing adverse impact on cattle production.

In this study, we implement Particle Filter (PF)-based assimilation algorithms to improve root-zone soil moisture (RZSM) estimates from a coupled SVAT-vegetation model during a growing season of sweet corn in North Central Florida. The results from four different PF algorithms were compared with those from the Ensemble Kalman Filter (EnKF) when near-surface soil moisture was assimilated every 3 days using both synthetic and field observations. In the synthetic case, the PF algorithm with the best performance used residual resampling of the states and obtained resampled parameters from a uniform distribution and provided reductions of 76% in root mean square error (RMSE) over the openloop estimates. The EnKF provided the RZSM and parameter estimates that were closer to the truth than the PF with an 84% reduction in RMSE. When field observations were assimilated, the PF algorithm that maintained maximum parameter diversity offered the largest reduction of 16% in root mean square difference (RMSD) over the openloop estimates. Minimal differences were observed in the overall performance of the EnKF and PF using field observations since errors in model physics affected both the filters in a similar manner, with maximum reductions in RMSD compared to the openloop during the mid and reproductive stages.

A study was conducted to evaluate the ability of the CERES-Maize model in the Decision Support System for Agrotechnology Transfer (DSSAT) to simulate sweet corn (Zea mays L. var. saccharata) yield and nitrogen leaching in Florida, considering input parameter uncertainties. In this type of biological system modeling, uncertainties in predictions with respect to input parameter uncertainty are often not reported. Thus, the result of model verification could be misleading if there are large uncertainties in field observations, since single model prediction values cannot comprehensively represent heterogeneous field conditions. Instead, comparisons between the distributions of model simulations and field observations were recommended in this study. A two-factor split-plot field experiment was conducted with three nitrogen fertilizer levels (185, 247, and 309 kg N ha-1) and two irrigation levels (I1 and I2; I2 = 1.5 × I1, where I1 is the irrigation demand calculated based on a daily soil water balance). Yield response to different nitrogen fertilizer and irrigation management levels was evaluated, and the cumulative nitrogen leaching was estimated for each of the treatments based on a nitrogen balance. Next, the field experiment treatments were simulated with the calibrated CERES-Maize model using parameter sets generated from parameter distributions derived with the generalized likelihood uncertainty estimation (GLUE) method in a previous study. Simulated dry matter yields and cumulative nitrogen leaching were compared to field-measured or estimated values. Measured total and marketable yields were not affected by irrigation level. Estimated nitrogen leaching increased significantly with higher levels of irrigation and nitrogen fertilizer application. The calibrated CERES-Maize model accurately predicted the phenology dates, with an error of 0 and 1 day for anthesis and maturity dates, respectively. The prediction uncertainties (due to uncertain input parameter values), as measured by the standard deviation (SD) in predicted anthesis and maturity dates, were only 1 and 2 days after planting, respectively. The model also accurately predicted the changes in dry matter yield caused by different nitrogen and irrigation levels, with a relative absolute error (RAE) less than 12% for all but one treatment. Due to the uncertainties in soil and genetic parameters, the prediction SD of simulated dry yields ranged from 655 kg ha-1 at I1 to 960 kg ha-1 at I2, while the observation SD ranged from 220 to 463 kg ha-1 for measured dry yields. The uncertainties in simulated dry yield were higher than the uncertainties of measured values due to relatively high variations in estimated genetic coefficients. The model performance could be improved further if the variations in estimated genetic coefficients could be reduced. The difference between the simulated and estimated nitrogen leaching amounts was significant and complex, ranging from -31 to 43 kg N ha-1 with an average absolute difference of 15.3%. This discrepancy was probably due to both the errors in estimation of potential nitrogen leaching in the field experiment using a mass balance approach and the inaccuracy of model predictions. Nevertheless, the increase in nitrogen leaching resulting from higher nitrogen fertilizer levels was correctly predicted. The uncertainties in simulated N leaching covered more than 67% of the uncertainties of estimated leaching for all but one treatment, indicating that estimated soil parameters via the GLUE method were able to represent the heterogeneity of field soil. In general, the CERES-Maize model is able to simulate sweet corn production under different management conditions sufficiently to allow exploration of tradeoffs between crop yield and nitrogen leaching for sweet corn production in Florida.

Flow of water through stream networks directly impacts flooding and transport of sediments and pollutants in watershed systems. Hence, knowledge of streamflow is critical for water management and mitigation of flooding and drought events. Unfortunately, spatially dense networks of in situ streamflow measurements are generally unavailable and would be prohibitively expensive to deploy and maintain. Thus, a data fusion framework is needed that utilizes available data to predict streamflow. Observed data in spatial (e.g., topography and land cover), temporal (e.g., streamflow and groundwater levels), and spatiotemporal domains (e.g., rainfall) impact streamflow. Some of these quantities can be obtained from remote sensing imagery; however, combining such disparate data types using traditional data fusion methods is problematic. Physically based hydrologic models have been used to predict streamflow but often with significant uncertainty because numerous assumptions are made for many unmeasured input and parameter values. Traditional Bayesian inference approaches suffer from superlinear increases in computational complexity as the number of data sets to be fused grows. In this paper, a scalable spatiotemporal approach based on Bayesian networks (BNs) is presented for estimating streamflow. An information-theoretic methodology based on conditional entropy is employed to quantify the impact of adding nodes in the BN in terms of information gained. The framework offers the flexibility of embedding knowledge from hydrologic models calibrated for the study area by introducing them as additional nodes in the network, thereby improving prediction accuracy. Posterior probabilities of estimates and the associated entropy provide valuable information on the quality of predictions and also offer directions for future watershed instrumentation.

Proper estimation of model parameters is required for ensuring accurate model predictions and good model-based decisions. The generalized likelihood uncertainty estimation (GLUE) method is a Bayesian Monte Carlo parameter estimation technique that makes use of a likelihood function to measure the closeness-of-fit of modeled and observed data. Various likelihood functions and methods of combining likelihood values have been used in previous studies. This research was conducted to determine the effects of using previously reported likelihood functions in a GLUE procedure for estimating parameters in a widely-used crop simulation model. A factorial computer experiment was conducted with synthetic measurement data to compare four likelihood functions and three methods of combining likelihood values using the CERES-Maize model of the Decision Support System for Agrotechnology Transfer (DSSAT). The procedure used an arbitrarily-selected parameter set as the known ‘‘true parameter set” and the CERES-Maize model to generate true output values. Then synthetic observations of crop variables were randomly generated (four replicates) by using the simulated true output values (dry yield, anthesis date, maturity date, leaf nitrogen concentration, soil nitrate concentration, and soil moisture) and adding a random observation error based on the variances of corresponding field measurements. The environmental conditions were obtained from a sweet corn (Zea mays L.) experiment conducted in 2005 in northern Florida. Results showed that the method of combining likelihood values had a strong influence on parameter estimates. The combination method based on the product of the likelihoods associated with each set of observations reduced the uncertainties in posterior distributions of parameter estimates most significantly. It was also found that the likelihood function based on Gaussian probability density function was the best among those tested. This combination accurately estimated the true parameter values, suggesting that it can be used when estimating CERES-Maize model parameters for real experiments.

In this study, an EnKF-based assimilation algorithm was implemented to estimate root-zone soil moisture (RZSM) using the coupled LSP–DSSAT model during a growing season of corn. Experiments using both synthetic and field observations were conducted to understand effects of simultaneous state–parameter estimation, spatial and temporal update frequency, and forcing uncertainties on RZSM estimates. Estimating the state–parameters simultaneously with every 3-day assimilation of volumetric soil moisture (VSM) observations at 5 depths lowered the average standard deviation (ASD) and the root mean square error (RMSE) for RZSM by approximately 1.77% VSM (78%) and 2.18% VSM (93%), respectively, compared to the open-loop ASD where as estimating only states lowered the ASD by approximately 1.26% VSM (56%) and the RMSE by 1.66% VSM (71%). The synthetic case obtained RZSM estimates closer to the observations than the MicroWEX-2 case, particularly after precipitation/irrigation events. The differences in EnKF performance between MicroWEX-2 and synthetic observations may indicate other sources of errors in addition to those in parameters and forcings, such as errors in model biophysics.

This paper examines the influence of military land use parameters on dimensionless indices related to storm flow, baseflow, and precipitation for five watersheds with areas ranging from 0.76 to 25.01 km2 within the Fort Benning Military Installation in southeastern US. Average magnitude and variability of these indices are grouped into four key hydrologic regimes—magnitude, frequency, duration, and rate of change. Correlation and regression analyses were performed to determine the relationship among the watershed physical characteristics and indices. A number of significant relationships were found. The correlation results show that the increase in road density increased the variability in the peak discharges and the slopes of the rising limb. The increase in the military training land increased the variability in the time base. The number of roads crossing streams is positively correlated with the response lag. Stepwise multiple correlations showed that the storm-based magnitude and variability in peak discharge, baseflow index, and the bankfull discharge have been significantly affected by military management related watershed characteristics. The relationship among the watershed physical characteristics and the storm-based hydrologic indices indicated that the greatest impact of land management is found with statistically significant predictive models for indices of time base, response lag, and time of rise. Military training land, road density, and the number of roads crossing streams were the three management variables that impacted storm responses.

Participation in carbon (C) markets could provide farmers in developing countries incentives for improving soil fertility. However carbon traders need assurances that contract levels of C are being achieved. Thus, methods are needed to monitor and verify soil C changes over time and space to determine whether target levels of C storage are being met. Because direct measurement over the large areas needed to sequester contract amounts of C in soil is not practical, other approaches are necessary. An integrated approach is described in which an Ensemble Kalman Filter (EnKF) is used to assimilate in situ soil carbon measurements into a stochastic soil C model to estimate soil C changes over time and space. The approach takes into account errors in in situ measurements and uncertainties in the model to estimate mean and variance of soil C for each land unit within a larger land area. The approach requires initial estimates of soil C over space along with uncertainties in these estimates. Model predictions are made to estimate soil C for the next year, in situ soil C measurements update these predictions using maximum likelihood methods, and the spatial pattern of soil C mean, variance, and covariance thus evolve over time. This approach can also be used to provide yearly estimates of the changes in soil C over multiple fields, the variance in those estimates, and aggregate soil carbon mean and variance values each year. In this paper, the use of the EnKF is shown for an area in Ghana with 12 fields, comparing numbers of fields sampled each year and ways of selecting which fields to sample each year. The model predicts soil C changes over time using first order decomposition of existing soil C and addition of C from plant residues. The lowest intensity sampling method (sampling only 1/4 of the fields per year) resulted in the highest level of uncertainty in aggregate soil C estimate. Rotating sample fields each year improved the performance of the EnKF. These results demonstrated a quantifiable tradeoff between field sampling intensity and uncertainty in aggregate soil C estimates. The framework could be modified to use more complex biophysical models and to assimilate remote sensing data

Among various ways to sequester CO2 from the atmosphere, increasing soil carbon is an option that could also lead to increased agricultural productivity, especially in developing countries. In order to accept this option as a mechanism for reducing atmospheric CO2 levels, a reliable soil carbon accounting system needs to be developed. Soil carbon can be estimated from direct measurements, but measurements often have higher variability than annual soil carbon changes due to the uncertain nature of in-situ sampling methods and laboratory analysis techniques. To reduce such variability in soil carbon estimates, the estimation accuracy of an approach that assimilates measurements with simulated outputs from a complex crop model was studied. The DSSAT-CENTURY model and the ensemble Kalman filtering (EnKF) method were used to test if this approach can provide more reliable estimates of SOC over time than measurements alone. The EnKF was developed to estimate soil carbon, crop biomass, and a model parameter that specifies the rate of soil organic matter decomposition. A base-case simulation study was set up with a continuous maize farming system in Ghana for 20 years. Based on an identical twin test that used model-generated truth and measurement datasets, the EnKF estimates of soil carbon showed superior estimation accuracy to measurements with less uncertainty. In time, the uncertainty in soil carbon estimates based on annual sampling was reduced by about 60% using the EnKF method. The method was not effective in estimating the model carbon decomposition rate parameter, mainly because of the low correlations of that parameter with either soil carbon or crop biomass. However, even when uncertainties in measurements and model predictions were relatively high, estimates of soil carbon converged to the true value, although uncertainties in those estimates varied with measurement and model uncertainties. We conclude that the use of the EnKF method with the complex DSSAT-CENTURY model would provide more reliable soil monitoring system, but more research is needed to compare complex versus simple models in this approach.

Rapid, repeatable assessment of ecological condition is critical for quantitative ecosystem monitoring. Soils provide a sensitive, integrative indicator for which sampling and analysis techniques are well defined. We evaluated soil properties as indicators of ecological condition (subjectively classified into minimally/moderately/severely degraded based on vegetative, hydrologic and edaphic cues) at 526 sites within Ft. Benning military installation (Georgia, USA). For each sample, we measured 17 biogeochemical parameters, and collected high-resolution diffuse reflectance spectra using visible/near infrared reflectance spectroscopy (VNIRS). VNIR spectra have been related to numerous soil attributes - we examine them here for diagnosing integrated response (i.e., ecological condition). We used ordinal logistic regression (OLR) and classification trees (CT) to discriminate between condition categories using both sets of predictors (biogeochemistry and spectra). Sixteen biogeochemical parameters were significantly different across condition categories; however, multivariate models greatly improved discrimination ([calibration, validation] accuracy of [69%, 66%] and [96%, 73%] for OLT and CT models, respectively). Important predictors included total C, total P, and Mehlich K/Ca/Mg. VNIR spectra further improved discrimination ([calibration, validation] accuracy of [74%, 70%] and [96%, 75%] for OLR and CT models, respectively). While spectra were comparably effective at discriminating minimally degraded sites, they were significantly more effective at discriminating severely degraded sites. Error rates across confounding factors suggest that watershed of origin and landscape position were the only important confounders, likely due to imbalanced sampling. We conclude that multivariate diagnosis improves accuracy, and that VNIR spectroscopy, which yields substantial cost and logistical improvements over conventional analyses, provides an effective tool for rapid condition diagnosis.

The literature to date on inverse modeling-based parameterization of spatial ("distributed ") crop models using yield data is dominated by one-dimensional uncoupled models, in which the landscape is divided into cells that do not interchange water. We studied three possible sources of error for uncoupled and spatially coupled distributed crop models: (1) error from biases (towards unrepresentatively wet or dry weather) in the sampling of years of yield data used for the parameterization process, (2) errors due to lack of knowledge of initial soil water conditions, and (3) error from lack of spatial coupling and water transport among different landscape locations. We show analytical evidence that the spatiotemporal infiltration behavior of a simple spatially coupled water balance model cannot be reproduced through a modification of the parameters of an uncoupled model. The corresponding yield prediction limitations of the uncoupled model are confirmed, using an example, both at the parameter estimation and the validation stages. In our example, however, parameter error due to weather biases and the error from lack of knowledge of initial conditions greatly impacted the predictive capability of the spatially coupled model, and had less effect on its uncoupled counterpart. This effect of biased weather has not been reported previously in the crop modeling IM literature. We conclude that the use of spatially coupled distributed crop models requires high-quality data. Practical precision agriculture applications are characterized by uncertain initial conditions and the possibility of biased weather. Under these circumstances, the use of a spatially coupled model may not be justified, especially for low landscape positions.

A physically based, variably saturated flow model was developed to predict soil water dynamics, evapotranspiration (ET) from the vadose zone, and recharge to (or exfiltration from) the saturated zone using mean daily atmospheric forcings and to identify the value of diurnal climate forcings on those predictions. The vadose zone flow is modelled using the Galerkin finite element technique to solve Richards’ equation in one-dimension. The model was able to accurately predict measured soil moisture, water table elevation and actual ET at Paynes Prairie State Preserve in Florida. The forecast Nash–Sutcliffe efficiencies of actual ET, water table and average soil moisture content increased modestly, from 0.605–0.653, 0.888–0.916 to 0.902–0.913, respectively, when the average daily ET forcing was replaced with a diurnal evaporation cycle. Several additional numerical experiments were conducted to evaluate the influence of the evaporation cycle disaggregation approach on modelled ET and soil moisture content for different soil textures, vegetation surfaces, and water table depth. The results show that the enhanced predictive value of the diurnal ET cycle increases with decreasing vegetation, decreasing clay content, and increasing water table depth. Using numerical studies, actual evaporation is shown to be higher for daily average evaporation as compared to the diurnal cycle evaporation for specific ranges of shallow water table depth. For clay soils, this range occurs from approximately 40 to 300 cm below land surface for bare soils and from approximately 40 to well below 300 cm below land surface for vegetated soils. The range for sandy soils is approximately 80–200 cm below land surface for both bare and vegetated soils. Within this range, the maximum difference of the actual to potential evapotranspiration ratio for the clay soil, resulting from using different forcing methods, is 20 and 10% for bare soil and vegetated conditions, respectively. The forcing method choice is more important for sandy soils with ratio differences of 50 and 16% for bare soil and vegetated conditions, respectively.

In this study, remotely sensed data and geographic information system (GIS) tools were used to estimate storm runoff response for Simms Creek watershed in the Etonia basin in northeast Florida. Land cover information from digital orthophoto quarter quadrangles (DOQQ), and enhanced thematic mapper plus (ETM+) were analyzed for the years 1990, 1995, and 2000. The corresponding infiltration excess runoff response of the study area was estimated using the U.S. Department of Agriculture (USDA), Natural Resources Conservation Service Curve Number (NRCS-CN) method. A digital elevation model (DEM)/GIS technique was developed to predict stream response to runoff events based on the travel time from each grid cell to the watershed outlet. A comparison of predicted to observed stream response shows that the model predicts the total runoff volume with an efficiency of 0.98, the peak flow rate at an efficiency of 0.85, and the full direct runoff hydrograph with an average efficiency of 0.65. The DEM/GIS travel time model can be used to predict the runoff response of ungaged watersheds and is useful for predicting runoff hydrographs resulting from proposed large scale changes in the land use.

Drinking water monitoring data have indicated anincrease in nitrate-nitrogen (NO3-N) concentration ingroundwater in some parts of the citrus production region ofFlorida. A proactive, incentive-based program of developingcrop-specific best management practices (BMP) began with theFlorida N-BMP legislation passed in 1994. A combination ofcareful irrigation and nitrogen (N) management is needed toimprove N uptake efficiency and to minimize potential leaching ofnitrate (NO3-N) to the groundwater. An improved Nmanagement practice is considered as a BMP, only if that practiceis proved to decrease NO3-N leaching into groundwater incommercial groves without adversely impacting the economics ofproduction. Therefore, long-term evaluation of horticulturalresponses as well as monitoring of groundwater NO3-N wereconducted in five commercial groves representing different soiltypes, citrus variety and rootstock, tree age, and culturalpractices to determine the impact of changes in N managementand/or irrigation scheduling. Groundwater NO3-N, leafnutrient concentrations, fruit yield and fruit quality weremonitored for 15 months under the growers' routine managementand, subsequently for 48 months, with improved N and irrigationmanagement practices. The N management practices evaluated inthis study included broadcast application of a combination ofwater soluble and slow release granular products, fertigation,and a combination of foliar application and fertigation. Irrigation management was improved by using tensiometer set pointof 10 and 15 cbar. This article presents the fruit yield, andconcentrations of N, P, K in six-month spring flush during thestudy period. The study showed that 5 to 8 yr old Valenciatrees on Volkamar lemon rootstock produced high quality fruit inthe range of 59 to 81 Mg ha-1 with 168 kg N ha-1 asfertigation combined with improved irrigation scheduling. Fruityield of 36 yr old Valencia orange trees on Rough lemonrootstock was greater with application of 180 kg N ha-1 yr-1 as fertigation compared to that of the trees whichreceived a similar rate of N as three broadcast applications ofgranular product. Fertilizer program comprising three foliarapplications of N using low biuret urea to deliver 66 kg N ha-1 yr-1 and an additional 76 kg N ha-1yr-1as fertigation was the most effective for decreasing the surficialgroundwater NO3-N while maintaining optimal fruit productionand nutritional status of the leaves. This study demonstratedthat economically and technically feasible N-BMPs can bedeveloped for citrus grown on sandy soils with a combination ofimproved N management and irrigation scheduling.

Groundwater recharge from the unsaturated zone results in a nonuniform mean groundwater flow. We report closed-form expressions for the mean and variance of advective solute travel time under such nonuniform flow conditions. These expressions are derived directly from velocity statistics, based on an analytical direct evaluation method (DEM). An alternative, indirect evaluation method (IEM) has previously been used for deriving and evaluating the cumulative distribution function of solute travel time in nonuniform mean flow, through its relation to an assumed Gaussian probability density function of solute displacement. For negative groundwater recharge we argue that neither of the two analytical methods DEM and IEM can be considered reliable. For positive groundwater recharge we show that (1) the DEM results reported here remain as reliable for nonuniform as for uniform mean groundwater flow; (2) an assumed lognormal cumulative distribution function based on the DEM-derived travel time moments is consistent with the IEM-derived solute travel time cumulative distribution function (cdf), which is in turn consistent with numerical simulation results reported in the literature; and (3) the detailed differences between the DEM-based lognormal cdf and the IEM-derived cdf for nonuniform flow are such that the former (DEM) may exhibit even better agreement with reported numerical simulation results than the latter (IEM).

The transport of nonreactive solutes through a coupled, two-dimensional, randomly heterogeneous vadose-saturated zone system subject to temporally random rainfall is predicted by Monte Carlo simulation and compared with previously published analytic results for three different rainfall patterns. The relative contributions of the uncertain inputs (i.e., rainfall and saturated conductivity) to the prediction uncertainty of solute transport are quantified in terms of the statistical moments of the pore water velocity, the plume spatial moments, and solute flux breakthrough curves at downstream control planes. Results show that the mean and variance of the saturated zone pore water velocity were approximately equivalent for the cases of uniform and random rainfall and were well predicted by the analytical relationships developed by Rubin and Bellin [1994]. As a result, the mean plume displacement, estimated by the trajectory of the mean plume center of mass, was found to be nearly identical for these cases. In the temporally random rainfall case, the saturated zone mean plume experienced more spread in the direction of mean flow at early times. However, the asymptotic rates of spatial spreading of the mean solute plumes were found to be approximately equivalent for the uniform and random rainfall cases and well predicted by the approximate expressions for longitudinal macrodispersivity in nonuniform flow proposed by Destouni and Graham [1995]. Random rainfall and random soil properties increased prediction uncertainty of the solute plume behavior in the vadose zone by an order of magnitude when compared with the uniform rain and random soil case. This effect was reduced considerably when the solute entered the saturated zone, where random rainfall produced only slightly larger prediction uncertainty than the uniform rainfall case. The analytic model developed by Destouni and Graham [1995] accurately predicted the temporal breakthrough of the mean solute plume at saturated zone control planes for all cases, if transport through the unsaturated zone accounted for the effects of temporally random rainfall using the methodology developed by Foussereau et al. [2000a, 2000b]. Results of this work indicate that for the humid climates studied here, uncertain rainfall patterns dominate transport prediction uncertainty in the shallow unsaturated zone, while uncertain solute breakthrough to the saturated zone and uncertain hydraulic conductivity dominate prediction uncertainty in the saturated zone.