Agricultural Water Management 132 (2014) 37–45 Contents lists available at ScienceDirect Agricultural Water Management j ourna l h omepage: www.elsevier.com/locate/agwat Drought monitoring using a Soil Wetness Deficit Index (SWDI) derived from MODIS satellite data Mohammad Reza Keshavarz a,∗ , Majid Vazifedoust b , Amin Alizadeh a a College of Agriculture, Ferdowsi University of Mashhad, Iran b College of Agriculture, Guilan University, Guilan, Iran a r t i c l e i n f o Article history: Received 26 April 2013 Accepted 10 October 2013 Keywords: Drought Soil wetness SWDI Remote sensing a b s t r a c t Soil moisture is considered a key index of agricultural drought monitoring systems due to its importance for plant growth and biological interactions. In this research, a Soil Wetness Deficit Index (SWDI) was developed based on a Soil Wetness Index to evaluate soil moisture deviation as an indicator of agricultural drought. The Soil Wetness Index is derived using a triangle space concept between the land surface temperature (LST) and vegetation index (NDVI). To acquire the triangle space concept, 8-day-products of land surface reflectance and LST derived from MODIS satellite data over Isfahan were used. The data was collected in the period of 2000–01 (dry year) and 2004–05 (wet year) on an 8-day time step. The results indicated that the SWDI index has the capability of mapping the spatial distribution of areas affected by drought, as well as the drought intensity. The estimated cumulative number of dry days (with -4 < SWDI < 0) in the period of 2000–01 was 184 days. The results also confirmed the existence of wet days in the period 2004–05. Moreover, shifts in drought condition at the end of the wet and dry periods were detected in the area. Results also showed that the presence of vegetation plays an important role in balancing soil moisture variation. © 2013 Elsevier B.V. All rights reserved. 1. Introduction Drought is the result of climate variations that infrequently occur in vast geographic areas without any specific border. Drought affects food security more than any other natural disaster. Predict- ing when a drought will occur and the length of its duration is very difficult compared to other disasters such as flood flashing (Kogan, 1997). Based on the environmental sub-system it affects, droughts are classified into meteorological, soil moisture (agricul- tural), hydrological and famine (Peters, 2003). Soil moisture content plays an effective role in agricultural drought (Kuenzer et al., 2008). Root zone moisture content is com- monly assigned to the upper 1–2 m of soil profile. This moisture is generally available for crop growth and can be transported to the atmosphere through the evapo-transpiration process (Verstraeten, 2006). Considering the effect of soil moisture on plant growth and crop productions, estimating soil moisture content is very impor- tant in monitoring agricultural drought. Nonetheless, due to the extreme variation of soil moisture along a day, it cannot be an appropriate index of drought severity by itself. Hence, an index is needed to consider the long-term variation of the soil moisture. ∗ Corresponding author. Tel.: +98 9126514204. E-mail address: mohammadreza.4231@gmail.com (M.R. Keshavarz). Also, the index should be able to compare the soil moisture deficit across seasons. Principally, soil moisture monitoring using field scale methods is very expensive, time consuming and impossible for application in vast areas. Additionally, converting field scale soil moisture data to geo-spatial maps using geo-statistical methods have no adequate quality. Therefore, there is a need for special tools that monitor spatial and temporal soil moisture variations continuously and accurately. Due to higher spatial and temporal resolution, methods based on remote sensing techniques in contrast to field mea- surements and simulation models are preferred for the regional purposes. Over the past two decades, numerous approaches have been developed for regional estimation of soil moisture. Some of these have been based on remotely sensed optical and microwave data. Methods include the use reflective data (Peters et al., 1991; Wang et al., 2007), thermal infrared data (Crow et al., 2008; Gillies and Carlson, 1995; Qui, 2006), passive and active microwave data (De Ridder, 2000; Mattia et al., 2008; Moran et al., 2004). Due to higher scatter in interaction with aerosols and lower penetration into the surface, optical and thermal imagery is more limited in com- parison to radar and microwave imagery. However, because of high spatial and temporal resolution and high correlation between soil moisture and surface temperature, implementation of optical and thermal imagery methods has increased in the recent years (Verstraeten, 2006). 0378-3774/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.agwat.2013.10.004