 2017 AARS, All rights reserved. * Corresponding author: jbakri@ju.edu.jo, jawad.t.albakri@gmail. com, Phone: +962-6-5355000-22444; Fax: +962-6-5300806 Assessment of Remote Sensing Indices for Drought Monitoring in Jordan Jawad T. Al-Bakri 1* , Areej Al-Khreisat 1 , Sari Shawash 2 , Eman Qaryouti 1 and Muna Saba 3 1 Department of Land, Water and Environment, Faculty of Agriculture, The University of Jordan, Amman - Jordan 2 Projects Directorate, The Hashemite Fund for Development of Jordan Badia, Amman – Jordan 3 Drought Monitoring Unit, National Center for Agricultural Research and Extension, Baqa’a, Jordan Abstract Remote sensing has been widely used in monitoring vegetation and in detecting agricultural droughts. The most commonly used data for this purpose is the coarse spatial and the high temporal resolution data of NDVI. This study compares different indices (NDVI, MSI, EWSI, PDI and MPDI) derived from MODIS for assessing drought conditions in Mafraq area in Jordan. The possible improvements in drought monitoring as a result of improved spatial resolution is also investigated in this study by comparing Landsat-NDVI with MODIS-NDVI. Results showed significant relationships among the different indices derived from MODIS and Landsat data. A significant relationship was found between Landsat-NDVI and MODIS-NDVI, with R 2 value of 0.56 and RMSE of 0.078. The Landsat-NDVI was better than MODIS-NDVI in detecting drought conditions for fields of rainfed barley in the northern parts of the study area, while both datasets reflected the aridity conditions prevailing in the study area. The MODIS-PDI showed to be the most accurate indicator that was highly correlated (R 2 = 0.73) with soil moisture measurements. Soil water stress indicators (EWSI and MSI) showed relatively lower correlations with soil moisture and modeled evapotranspiration when compared with Landsat-NDVI and MODIS-PDI. Therefore, the use of MODIS-PDI instead of MODIS- NDVI would be recommended for mapping drought severity without processing historical data. The use of NDVI deviations from historical means would be recommended with medium resolution data of NDVI, providing that temporal resolution would improve and more datasets from earth observation systems (EOS) would be available in real time. Key words: Drought, remote sensing, NDVI, MPDI, SEBAL-ETa, Landsat, MODIS, Jordan. 1. Introduction Drought is a natural phenomenon that is related to reduction in rainfall amounts received over an extended period, such as a season or a year, resulting in insufficient moisture stored in the soil (McKee et al., 1993). Different types of droughts can be defined, including the meteorological, agricultural and hydrologic types. The meteorological type of drought can be categorized for different time intervals, while agricultural drought has typically a short-time scale of one month when soil moisture and rainfall are inadequate to support crop growth leading to the loss of yield. The hydrological droughts have intermediate and long-time scales of 3, 6 and 12 months or higher, with marked depletion of surface and subsurface water (Wilhite and Glantz, 1985; Wilhite, 2000). Regardless of the type of drought, its frequencies and severity have increased and resulted in increasing the areas affected by this adverse phenomenon, particularly in arid environments. This is mainly attributed to climate change that resulted in changing meteorological characters such as temperatures; winds; relative humidity; and rainfall patterns and amounts (IPCC, 2007; Mishra and Singh, 2010). Subsequently, freshwater availability and agricultural production started to be adversely impacted by drought, leading to economic losses in many sectors, especially agriculture. Thus, mapping of drought is important and needed for assessment of its impacts on the different sectors