An investigation of drought prediction using various remote-sensing vegetation indices for different time spans. Issue 6 (19th March 2018)

• Record Type: Journal Article
• Authors: Heydari, Hamed; Valadan Zoej, MohammadJavad; Maghsoudi, Yasser; Dehnavi, Sahar
• Is Part Of: International journal of remote sensing. Volume 39:Issue 6(2018)
• Page Start: 1871
• Page End: 1889
• Publication Date: 2018-03-19
• Subjects: Remote sensing -- Periodicals; Télédétection -- Périodiques; 621.3678
• DOI: 10.1080/01431161.2017.1416696
• Languages: English
• ISSNs: 0143-1161

Abstract

ABSTRACT: Iran is a country in a dry part of the world and extensively suffers from drought. Drought is a natural and repeatable phenomenon definable at specified time and area. In addition, social and economic issues can be affected by drought. Information such as intensity, duration, and spatial coverage of drought can help decision makers to reduce the vulnerability of the drought-affected areas, therefore lessen the risks associated with drought episodes. Lack of long-term meteorological data for many parts of the country is one of the most important problems for drought monitoring in Iran. One of the useful ways for gathering information about soil and vegetation conditions is using satellite-based imagery. In this study, remotely sensed image data were applied in order to forecast and model the drought. To this end, SPI (standardized precipitation index) drought indicator was used to represent the drought and its intensity in different time spans (1, 3, 6, 9, 12, and 24 months). Some vegetation indices (VIs) including normalized difference vegetation index, temperature condition index, vegetation condition index, and normalized difference vegetation index deviation were extracted using Advanced Very High Resolution Radiometer sensor imagery. These indices were plugged into the model to calculate the SPI. A unique Support Vector Machine classifier improved for all types of the SPI by applying various remotely sensed VIs. The best vegetation index for each kind of SPI was determined. In this framework, meteorological stations were clustered based on their land cover extracted from satellite-based indices before insertion to the model.