Integrating Satellite Evapotranspiration and Land Cover Distribution Maps to Assess the Spatiotemporal Impacts of Drought in Southwestern Colorado

Evapotranspiration (ET) is a key hydrologic variable that represents the response of the landscape to a combination of climatic, hydrologic and agronomic drivers. In this study, we investigated the sensitivity of different land covers to drought using percent reduction in landscape ET in southwestern Colorado. Seventeen years (2000-2016) of seasonal total (April to October), 1 km resolution ET dataset was generated using the Operational Simplified Surface Energy Balance (SSEBop) Model was intersected with a National Land Cover Dataset (NLCD 2011) to evaluate the spatiotemporal dynamics of ET. This study showed large differences in ET among different land cover classes, ranging from a low of less than 300 mm for barren lands to more than 600 mm for forested and wetland land cover classes. A time series analysis of 15 land cover classes showed comparable year-to-year temporal patterns, with a major reduction during known drought years, such as 2002 and 2012. In 2002, the ET reduction from the average of 17 years (2000-2016) was greater than 30% for all the land cover types except for mixed forest whose ET only reduced by 23%. In the same year, developed/high intensity land cover type had the highest percent reduction at 59%. Comparatively, for 2012 the overall ET reduction was less than 25% for all land cover types. Shrub/scrub cover type had the highest percent reduction at 25%, and open water had the least percent reduction at 12%. Using percent ET reduction as a metrics, 2002 was a more severe drought than 2012 by about a factor of 2 for most land cover types. Future work will expand the analysis to a larger region in Colorado and the U.S. to evaluate the relative sensitivity of different land covers under varying drought levels. Understanding the relative sensitivity of the landscape to drought is important for drought monitoring and fire risk analysis and management.