No abstract

No abstract

Droughts occurring every year all over the world have great impacts on human society, nature, and the global economy for example in declining crop yields, reduction of water supplies, and distressed vegetation. Satellite data have been widely used in drought monitoring. Vegetation condition is an excellent indicator of agricultural drought and can be quantified by the Normalized Difference Vegetation Index (NDVI). One way to detect agricultural drought is to quantify it through calculation of drought indices, such as the Vegetation Condition Index (VCI). For this purpose, we have developed an agricultural drought portal under the name of Global Agricultural Drought Information Services System (GADISS), which is interoperable with the Global Earth Observation System of Systems (GOESS) and follows Web standards for geospatial data recommended by the Open Geospatial Consortium (OGC), on the top of remote sensing drought monitoring. This study investigates the performance of VCI drought maps, which is also scientific base of GADISS, against annual grape production. Taking severe drought in the Aegean region, Turkey in 2007 as an example, the VCI drought index derived from 8-day NDVI satisfactorily detect drought. It is also validated with the fluctuations of annual grape production capacity. The results suggest that agricultural droughts have negative impacts on grape production rate per tree, and they can be operationally monitored by VCI over a geographic region for an extended period of time.

During the last decade, the use of the normalized difference vegetation index (NDVI) for drought monitoring applications has drawn many criticisms, mainly because a number of drivers such as land-cover/land-use change, pest infestation, and flooding may depress the NDVI, further causing false drought identification. In this study, the impacts of land-cover change on the NDVI-derived satellite drought indicator, the vegetation condition index (VCI), are presented. It was found that the VCI is sensitive to changes in land cover, especially deforestation, the land cover changes from evergreen and deciduous forests to other land-cover classes. However, because the scale of land-cover changes was very small across the study area, only trivial drought alerts were observed in the VCI-based drought maps during non-drought years. Because drought is a large-scale climate event, it is reasonable to neglect these alerts. Besides, when the VCI was averaged to climate division scale, the results obtained through the VCI method were in good agreement with those acquired by the meteorological data-based drought indices such as the Palmer drought severity index and standardized precipitation index.

No abstract