Agricultural Drought Monitoring and Analysis Using Remotely Sensed Data from NOAA-AVHRR

Drought is one of the most harmful environmental phenomena which has a devastating impact on agriculture, environment, economy and society. Drought happens during a period of water deficit in a region due to low precipitation, high evapotranspiration, high groundwater derivation or a mixture of the mentioned factors. Observations from space, especially from the NOAA polar-orbiting satellites, provide a synoptic view, permanent data archive, extra visual information, cost effectiveness, and a regular, repetitive view of nearly all of the earth’s surfaces (Johnson et al. 1993). Hence, remotely sensed data from NOAA’s AVHRR has been commonly used for national to global-scale drought monitoring activities since the 1980s and application of satellite based drought indices of vegetation health (TCI, VCI, VHI) in drought identifying, analyzing, evaluating and monitoring has been studied world widely (Kogan 1997, Vogt et al. 1998, Seiler et al. 2000, Singh et al. 2003, Kogan et al. 2004, Bhuiyan 2008, Rahimzadeh Bajgiran et al. 2008). Materials and methodsThe research was carried out in Isfahan Province which covers an area of about 105'937 km2 located between to and to. In this research, the precipitation data of 14 synoptic stations of Isfahan Province has been first studied and 4 synoptic stations (Isfahan, Shahreza, Natanz and Daran) have been chosen due to their long-term data record and good distribution across the area.In addition to field data, GVI-x data set is used which is provided by NOAA/NESDIS web site (http://www.star.nesdis.noaa.gov/smcd/emb/vci/VH/vh_browse.php). This data set contains NDVI, BT, TCI, VCI, VHI products that are put weekly as HDF files on the website. The used data have a weekly temporal resolution and a 16km2 spatial resolution and cover an area between to and to. Results and discussionMonthly vegetation health indices (TCI, VCI and VHI) in Isfahan Province have been computed from 1993 to 2007 using GVI-x weekly data set to study the application of these indices in drought monitoring. Then correlation of monthly precipitation and monthly air temperature of the selected synoptic stations with the amount of monthly TCI, VCI and VHI in bare, irrigated and rainfed locations (Derived from Isfahan Province land use map) have been calculated.The results indicate that the correlation between precipitation and TCI, VCI and VHI is significant at 10%, 40% and 40% of studied statistics at 90% confidence level. The same amounts for temperature are 40%, 35% and 35%. These results show that VCI and VHI are more reliable for drought monitoring because of their higher correlation with precipitation comparing to TCI which is more correlated with temperature.Studying the annual precipitation variability of selected stations in Isfahan Province during the period 1993-2007 and comparing monthly precipitation to the 15-year average precipitation indicate that the selected stations have experienced in most months of the years 2000, 2004 and 2005 respectively drought, wet and normal conditions. Then drought intensity in irrigated and rainfed locations have been obtained using the amounts of monthly VCI and VHI and table of classification of drought intensity using vegetation health indices (TCI, VCI and VHI) (Bhuiyan 2008). Areal aspect of TCI, VCI and VHI across the province has also been analyzed and drought intensity has been studied based on the table by Bhuiyan (2008) using vegetation health indices (TCI, VCI and VHI) during 1993-2007. ConclusionIn this research the satellite-based drought indices (TCI, VCI and VHI) have been computed in Isfahan Province using GVI-x data set. The correlation of these indices with precipitation data has been studied to evaluate and analyze the application of these indices in drought evaluating and monitoring. The results indicate rather a good accordance of these indices with precipitation variability during the study period and significant correlation between monthly VCI and VHI and monthly precipitation. VCI and VHI variations during the years 2000, 2004 and 2005 shows that in most months of the year 2000, the amount of the indices is less (means less vegetation) than 2004 and the curve of the year 2005 is mainly located between the 2000 and 2004 curves.