Comparison Of Drought Monitoring Indices Derived From MODIS and CHIRPS Data Using Google Earth Engine

Creative Commons License

Aksoy S. , Sertel E.

9th Global Conference on Global Warming (GCGW-2021), Zagreb, Croatia, 1 - 04 August 2021, pp.1-5

  • Publication Type: Conference Paper / Full Text
  • City: Zagreb
  • Country: Croatia
  • Page Numbers: pp.1-5


Drought is an important natural hazard causing adverse effects on the economy, agriculture, ecology, and human life. Increasing evapotranspiration as a result of higher temperature values and precipitation deficiency trigger drought conditions. Unlike the other natural hazards, drought is slowly developing disaster but its impacts are cumulative and destructive. It is foreseen that climate change will also impact the severity and frequency of droughts; therefore, it is crucial to monitor drought conditions and the impacts of these conditions on different sectors. Earth observation satellite data could be successfully used to monitor the spatio-temporal distribution of drought conditions from regional to continental scales. Numerous Remote Sensing indices on drought monitoring are available in the literature. These indices can provide information on the spatio-distribution of drought conditions as well as the severity of the drought.

Our main aim in this study is to comparatively analyze different drought indices derived from satellite images and meteorological drought indices to generate multi-temporal drought maps at the country level. We investigated drought conditions by applying NDDI, NMDI, and VHI MODIS satellite data. Meteorological drought indices namely PCI and SPI were created using CHIRPS data. VHI was calculated using VCI and TCI. VCI and TCI are calculated using multi-temporal satellite images to analyze the impact of temporal dimension on drought conditions. Drought severity maps were created using NDDI, NMDI, VHI, and SPI indices, and also time series graphs were plotted and compared with CHIRPS derived data having temporal resolution of 5 days. We produced monthly maps and time series due to the limitation caused by cloudiness and decrease in the availability of clear optical remote sensing data. Several cities in Turkey experiencing different climatic conditions were examined in this study. We implemented algorithms and produced maps and time series results using the GEE platform. GEE provides an extensive cloud-based environment to the users to rapidly analyze a huge number of earth observation data from different sensors without downloading them. GEE provides a significant storage area, various data analysis functions, and a possibility to utilize developed solutions in any other part of the world.