چكيده به لاتين
Abstract:
Drought is one of the natural disasters and recurring phenomena in all parts of the world with different climates and is not limited to arid and semi-arid areas. Our country is also involved with drought in different area. The catchment area of Lake Urmia is one of the most important watersheds in the country, which threatens the dangers of its drought phenomenon. Most drought calculation methods are based on rainfall data of ground stations, although these data are highly accurate, but have various disadvantages such as location constraints, low station densities, especially in Impassable points, and expensive cost of measurements at these stations. Therefore, the use of remote sensing systems in estimating rainfall in locations without a station seems to be necessary. In this study, using the data of ground stations in the catchment area of Lake Urmia during the years 1998 to 2016, the accuracy of satellite models in estimating precipitation and meteorological drought was evaluated. Various indicators for drought analysis have been presented that the SPI Meteorological Drought Index is the most widely used and more valid. The results indicate that both of the TRMM and GPCP models have average accuracy. The TRMM model in the estimation of SPI meteorological drought has a correlation coefficient of 0.75 and a error coefficient of RMSE is 0.7. Also, the GPCP model in SPI meteorological drought estimation has a correlation coefficient of 0.82 and a error coefficient of RMSE is 0.7 , which was better than other models. The PERSIANN-CDR model has a correlation coefficient of 0.65 and a RMSE error rate of 0.78. Although the PERSIANN model offers mediocre results in some parts of the catchment area, but in most parts of the model, it presents poor results. This model has a correlation coefficient of 0.45 and an error factor of RMSE is 1. Also, the CMORPH model did not allow for adequate estimation of drought and provided very poor results.
Keywords: Drought , PERSIANN، PERSIANN-CDR , TRMM ، CMORPH and GPCP.