Drought, one of the costliest natural disasters, lowers the productivity of grasslands, directly affecting the number of cattle that can be raised. Suitable drought monitoring tools that indicate the effect of drought on grassland productivity can help producers with their decision-making. Remotely sensed vegetation indices can be used to monitor the impacts of drought on vegetation health. However, it may be challenging to distinguish between weather and other vegetation health stressors (i.e. fire, overgrazing). This study develops a data-driven multiple regression grass productivity model that uses environmental and climate variables to separate the effect of weather from other disturbances. The satellite-derived Normalized Difference Vegetation Index (NDVI) is used as a proxy for grassland productivity and is converted to a total growing season biomass using a previously developed empirical equation. A biomass productivity map is generated over the Sandhills ecoregion using model results. Additionally, productivity extracted from 1000 randomly placed points in the Sandhills area is regressed on various drought indices and tools to evaluate which ones can better explain the interannual variability in productivity caused by a drought. The multiple indices and tools used in this study include the Standardized Precipitation Index (SPI) and the United States Drought Monitor.
Results/Conclusions
The Expected Ecosystem Performance (EEP) model was evaluated on a set of testing points (n = 1530, r = .93). EEP maps derived from the model adequately capture the wet and drought years as well as the productivity gradient observed in the Sandhills area. Seasonal biomass anomaly maps, based on the deviation of EEP from a long-term median, capture additional information about the spatial variability of growing season biomass based on the spatial distribution of drought conditions. This research also uniquely combines the EEP results with various drought indices to evaluate drought based on numerous input variables. The best performing drought indices in the regression analysis were 3- and 6-month SPI (r2= .71 and .70 respectively). These data, if provided on a near-real time basis would supply livestock producers with valuable information that can help land managers to better understand drought impacts on grassland systems and to make optimal management decisions based on predicted forage scenarios.Based on these results we will develop an interactive management decision tool with 250-meter spatial resolution and a large spatial extent to provide producers with an estimate of total growing season forage production under various weather scenarios.