Volume XLII-4/W18
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-4/W18, 749–753, 2019
https://doi.org/10.5194/isprs-archives-XLII-4-W18-749-2019
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-4/W18, 749–753, 2019
https://doi.org/10.5194/isprs-archives-XLII-4-W18-749-2019
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

  18 Oct 2019

18 Oct 2019

NEURAL NETWORK METHOD FOR DROUGHT MODELING USING SATELLITE DATA

R. Mokhtari and M. Akhoondzadeh R. Mokhtari and M. Akhoondzadeh
  • Remote Sensing Department, School of Surveying and Geospatial Engineering, College of Engineering,University of Tehran, North Amirabad Ave., Tehran, Iran

Keywords: Drought Model, Neural Network, MODIS, SMOS, TRMM, NDVI

Abstract. Drought is one of the natural crises in each region. Drought has a direct relationship with vegetation. Various factors affect vegetation. The relationship between these factors and vegetation can be expressed using methods of machine learning algorithms. Nowadays, using remote sensing images can be used to measure the factors affecting vegetation and investigate this phenomenon with high precision. In this research, vegetation and various factors affecting this factor, which can be measured using satellite imagery, are selected. The factors include land surface temperature (LST), evapotranspiration (ET), snow cover, rainfall, soil moisture that which are derived from the active and passive sensors of satellite sensors as the products of land surface temperature (LST), snow cover and vegetation, using images of products of the MODIS sensor and rainfall using the images of the TRMM satellite and soil moisture using the images of the SMOS satellite during a period from June 2010 to the end of 2018 for the central region of Iran has received and after that, primary processing was performed on these images. The vegetation index (NDVI) is modeled using artificial neural network algorithm for monthly periods. method have been able to achieve model with desirable accuracy. The average accuracy was RMSE = 0.048 and R2 = 0.867.