Search

The ecosystem performance approach, used in a previously published case study focusing on the Nebraska Sandhills, proved to minimize impacts of non-climatic factors (e.g., overgrazing, fire, pests) on the remotely-sensed signal of seasonal vegetation greenness resulting in a better attribution of its changes to climate variability. The current study validates the applicability of this approach for assessment of seasonal and interannual climate impacts on forage production in the western United States semi-arid grasslands. Using a piecewise regression tree model, we developed the Expected Ecosystem Performance (EEP), a proxy for annual forage production that reflects climatic influences while minimizing impacts of management and disturbances. The EEP model establishes relations between seasonal climate, site-specific growth potential, and long-term growth variability to capture changes in the growing season greenness measured via a time-integrated Normalized Difference Vegetation Index (NDVI) observed using a Moderate Resolution Imaging Spectroradiometer (MODIS). The resulting 19 years of EEP were converted to expected biomass (EB, kg ha−1 year−1) using a newly-developed relation with the Soil Survey Geographic Database range production data (R2 = 0.7). Results were compared to ground-observed biomass datasets collected by the U.S. Department of Agriculture and University of Nebraska-Lincoln (R2 = 0.67). This study illustrated that this approach is transferable to other semi-arid and arid grasslands and can be used for creating timely, post-season forage production assessments. When combined with seasonal climate predictions, it can provide within-season estimates of annual forage production that can serve as a basis for more informed adaptive decision making by livestock producers and land managers.

Invasive annual grasses, such as cheatgrass (Bromus tectorum L.), have proliferated in dryland ecosystems of the western United States, promoting increased fire activity and reduced biodiversity that can be detrimental to socio-environmental systems. Monitoring exotic annual grass cover and dynamics over large areas requires the use of remote sensing that can support early detection and rapid response initiatives. However, few studies have leveraged remote sensing technologies and computing frameworks capable of providing rangeland managers with maps of exotic annual grass cover at relatively high spatiotemporal resolutions and near real-time latencies. Here, we developed a system for automated mapping of invasive annual grass (%) cover using in situ observations, harmonized Landsat and Sentinel-2 (HLS) data, maps of biophysical variables, and machine learning techniques. A robust and automated cloud, cloud shadow, water, and snow/ice masking procedure (mean overall accuracy >81%) was implemented using time-series outlier detection and data mining techniques prior to spatiotemporal interpolation of HLS data via regression tree models (r = 0.94; mean absolute error (MAE) = 0.02). Weekly, cloud-free normalized difference vegetation index (NDVI) image composites (2016−2018) were used to construct a suite of spectral and phenological metrics (e.g., start and end of season dates), consistent with information derived from Moderate Resolution Image Spectroradiometer (MODIS) data. These metrics were incorporated into a data mining framework that accurately (r = 0.83; MAE = 11) modeled and mapped exotic annual grass (%) cover throughout dryland ecosystems in the western United States at a native, 30-m spatial resolution. Our results show that inclusion of weekly HLS time-series data and derived indicators improves our ability to map exotic annual grass cover, as compared to distribution models that use MODIS products or monthly, seasonal, or annual HLS composites as primary inputs. This research fills a critical gap in our ability to effectively assess, manage, and monitor drylands by providing a framework that allows for an accurate and timely depiction of land surface phenology and exotic annual grass cover at spatial and temporal resolutions that are meaningful to local resource managers.

The satellite-derived growing season time-integrated Normalized Difference Vegetation Index (GSN) has been used as a proxy for vegetation biomass productivity. The 250-m GSN data estimated from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensors have been used for terrestrial ecosystem modeling and monitoring. High temporal resolution with a wide range of wavelengths make the MODIS land surface products robust and reliable. The long-term 30-m Landsat data provide spatial detailed information for characterizing human-scale processes and have been used for land cover and land change studies. The main goal of this study is to combine 250-m MODIS GSN and 30-m Landsat observations to generate a quality-improved high spatial resolution (30-m) GSN database. A rule-based piecewise regression GSN model based on MODIS and Landsat data was developed. Results show a strong correlation between predicted GSN and actual GSN (r = 0.97, average error = 0.026). The most important Landsat variables in the GSN model are Normalized Difference Vegetation Indices (NDVIs) in May and August. The derived MODIS-Landsat-based 30-m GSN map provides biophysical information for moderate-scale ecological features. This multiple sensor study retains the detailed seasonal dynamic information captured by MODIS and leverages the high-resolution information from Landsat, which will be useful for regional ecosystem studies.

Land management practices and disturbances (e.g. overgrazing, fire) have substantial effects on grassland forage production. When using satellite remote sensing to monitor climate impacts, such as drought stress on annual forage production, minimizing land management practices and disturbance effects sends a clear climate signal to the productivity data. This study investigates the effect of this climate signal by: (1) providing spatial estimates of expected biomass under specific climate conditions, (2) determining which drought indices explain the majority of interannual variability in this biomass, and (3) developing a predictive model that estimates the annual biomass early in the growing season. To address objective 1, this study uses an established methodology to determine Expected Ecosystem Performance (EEP) in the Nebraska Sandhills, US, representing annual forage levels after accounting for non-climatic influences. Moderate Resolution Imaging Spectroradiometer (MODIS)-based Normalized Difference Vegetation Index (NDVI) data were used to approximate actual ecosystem performance. Seventeen years (2000−2016) of annual EEP was calculated using piecewise regression tree models of site potential and climate data. Expected biomass (EB), EEP converted to biomass in kg*ha−1*yr−1, was then used to examine the predictive capacity of several drought indices and the onset date of the growing season. Subsets of these indices were used to monitor and predict annual expected grassland biomass. Independent field-based biomass production data available from two Sandhills locations were used for validation of the EEP model. The EB was related to field-based biomass production (R2 = 0.66 and 0.57) and regional rangeland productivity statistics of the Soil Survey Geographic Database (SSURGO) dataset. The Evaporative Stress Index (ESI), the 3- and 6-month Standardized Precipitation Index (SPI), and the U.S. Drought Monitor (USDM), which represented moisture conditions during May, June and July, explained the majority of the interannual biomass variability in this grassland system (three-month ESI explained roughly 72% of the interannual biomass variability). A new model was developed to use drought indices from early in the growing season to predict the total EB for the whole growing season. This unique approach considers only climate-related drought signal on productivity. The capability to estimate annual EB by the end of May will potentially enable land managers to make informed decisions about stocking rates, hay purchase needs, and other management issues early in the season, minimizing their potential drought losses.

This study identifies areas with ecosystem performance anomalies (EPA) within the Upper Colorado River Basin (UCRB) during 2005–2007 using satellite observations, climate data, and ecosystem models. The final EPA maps with 250-m spatial resolution were categorized as normal performance, underperformance, and overperformance (observed performance relative to weather-based predictions) at the 90% level of confidence. The EPA maps were validated using “percentage of bare soil” ground observations. The validation results at locations with comparable site potential showed that regions identified as persistently underperforming (overperforming) tended to have a higher (lower) percentage of bare soil, suggesting that our preliminary EPA maps are reliable and agree with ground-based observations. The 3-year (2005–2007) persistent EPA map from this study provides the first quantitative evaluation of ecosystem performance anomalies within the UCRB and will help the Bureau of Land Management (BLM) identify potentially degraded lands. Results from this study can be used as a prototype by BLM and other land managers for making optimal land management decisions.

Regression tree models have been widely used for remote sensing-based ecosystem mapping. Improper use of the sample data (model training and testing data) may cause overfitting and underfitting effects in the model. The goal of this study is to develop an optimal sampling data usage strategy for any dataset and identify an appropriate number of rules in the regression tree model that will improve its accuracy and robustness. Landsat 8 data and Moderate-Resolution Imaging Spectroradiometer-scaled Normalized Difference Vegetation Index (NDVI) were used to develop regression tree models. A Python procedure was designed to generate random replications of model parameter options across a range of model development data sizes and rule number constraints. The mean absolute difference (MAD) between the predicted and actual NDVI (scaled NDVI, value from 0–200) and its variability across the different randomized replications were calculated to assess the accuracy and stability of the models. In our case study, a six-rule regression tree model developed from 80% of the sample data had the lowest MAD (MADtraining = 2.5 and MADtesting = 2.4), which was suggested as the optimal model. This study demonstrates how the training data and rule number selections impact model accuracy and provides important guidance for future remote-sensing-based ecosystem modeling.