As a part of repeating climate features in the world, drought is a natural hazard with a prolonged insufficiency in reasonable water supply beyond the range of normal human activities. Usually, drought is a slow-outbreak natural disaster of significant period and intensity over a noticeably large area. It is one of the most damaging and costly natural disasters potentially causing fatal results such as a wide and prolonged severe famine especially in the developing countries in Africa.

From Terra Moderate Resolution Imaging Spectroradiometer (MODIS), remotely sensed visible and infrared surface reflectance and surface temperature measurements are utilized to test long-term changes of vegetation drought conditions in the eastern Horn Of Africa (HOA). The MODIS reflective solar bands observations are utilized to measure well-known indexes called normalized difference vegetation index (NDVI) and normalized difference water index (NDWI). Because of the oscillations with these indices, a special polynomial window filtering is applied as a pre-processing step to detect time dependent changes in NDVI and NDWI. A linear regression of the filtered data shows that the NDWI has shown a non-negligible degradation, which indicates that on-going drought condition is getting worse in the eastern HOA.

The MODIS 8-day average Land Surface Temperature (LST) products also provide thermal condition changes in the eastern HOA region. Over the study period, the Terra MODIS daytime LST has negative slope approximately --0.037 degrees [K/year], but all other Aqua day and night results suggest positive slope values from 0.04 to 0.28 degrees [K/year]. The summation of the day and night temperatures from the two satellites shows positive slope approximately 0.025 degrees [K/year]. To identify spatial differences along with the time domain, the original dimension is binned to a smaller image size. The day LST change pattern shows that negative slopes in the northern hot and dry area and positive slopes are observed in the southern part of study area. The results also strongly suggest the eastern HOA is experiencing on-going warming during the last decade especially near the equator.

In the eastern HOA area, there is no ground weather station or a flux tower for ground truth to estimate drought situation. As an alternative choice, robust observations of a remotely sensed data sets such as NDVI and LST products are used to estimate hydrologic surface conditions. Combining the NDVI and LST data cloud in an orthogonal space, evapotranspiration (ET) is a key factor of the hydrological component, because it is calculated by the energy or moisture exchange between the atmosphere and the ground which consists of soil, water, and vegetation. Historically, ET has been utilized as a major indicator of drought, because not only it includes effects from the soil moisture content, but also anomalies of ET are related to the vegetation health and growing rate. In this study, a new histogram based Evaporative Fraction (EF) estimation method is developed and applied to estimate surface EF condition changes over the from Southern Great Plains (SGP) area in United States. A histogram based on an on-orbit EF estimation algorithm is developed to determine hot and cold edges in the triangle distribution using only remotely sensed data. The EF estimation algorithm is developed and compared with different options of existing hot and cold edge algorithms. The selected algorithms are applied to five-year Terra MODIS data products and the on-orbit EF estimations are validated using ground truth data from SGP. Out of numerous iterations, the 99.7 percent histogram based hot edge and the mean Fr 90 cold edge algorithm combinations provides an excellent performance between the ground truth and on-orbit EF estimations among the tested EF algorithms with the triangle method. Finally, the EF estimation method is applied to the NDVI and LST data over the eastern HOA area to detect time dependent severe drought condition changes through the spatial and temporal analyses. Averaged on-orbit EF response shows EF degradation over time with repeating extreme conditions observed during the second dry seasons in months of June, July, and August (JJA) annually. Consequently, the eastern HOA has been experiencing severe drought condition without any sign of improvement based on on-obit NDVI, NDWI, and EF observations especially in the second dry (JJA) and growing seasons.