Evaluation of accuracy for satellites rainfall datasets compared in ground stations: a case study of duhok governorate, Northern Iraq

Providing accurate and reliable rainfall data that can be used and applied in various climate and hydrological studies is essential. This paper aims to assess the accuracy of monthly rainfall data for satellites (climate engine, NASA-POWER) and corresponding data for ground stations, and through spatial mapping and linear measurement of rainfall indicators in Duhok Governorate in northern Iraq. The data evaluation process included the use of some statistical and cartographic methods available within the Jeffrey Amazing Statistics Program (JASP) and Origin Pro software, the evaluation and statistical analysis were conducted during the period from 2003 to 2022. The analysis results indicate that the relationship of ground stations with the climate engine recorded good values (Pbias = 1.24, NSE = 0.93, R = 0.97, Slope = 0.99, KGE = 0.72). However, these values are lower when compared with NASA-POWER (Pbias = 14.09, NSE = 0.55, R = 0.94, Slope = 0.34, KGE = -12.9). Both results indicate a positive relationship between the satellite and the ground data, and in general, all climate stations recorded a high correlation factor in monthly rainfall forecasting. Furthermore, the climate engine's data were characterized by high rainfall accuracy and quality and were relatively consistent with the observed scale data (ground stations). This study provides new ideas about the methods for selecting rainfall products for climatic and hydrological studies.