Spatio-Temporal Evaluation of Satellite Precipitation Products to Estimate Heavy Rainfall Events in Northwest Iran in the Last Two Decades

Extended Abstract

Introduction

Climate change has led to an increase in extreme events such as heavy rainfall. To prevent the damage of this atmospheric phenomenon, it is necessary to predict the patterns that cause it, which requires having heavy rainfall data in the past periods. Verification of different models to choose the optimal model for climate change issues requires having accurate data on heavy rainfall in the past decades. In mountainous areas, due to their topography, it is very expensive to build synoptic stations, and meteorological stations only provide researchers with point-by-point data. Satellites can compensate for these limitations with their universal ability to provide data as well as simple and fast access. Before using satellite products, it is necessary to make a proper assessment regarding the accuracy of different types of satellites and after verification, based on the atmospheric element and also the studied location, the best option should be selected. The purpose of this study is to evaluate the precipitation products of satellites in identifying the heavy rains in northwest Iran in the last two decades.



Materials and Methods

In this study, the precipitation data of 23 stations located in the northwest of Iran including the provinces of West Azarbaijan, East Azarbaijan, Ardabil, North Kurdistan, and West Zanjan were used in the period 2000-2019. First, the extreme occurrences of precipitation in the study area in the desired period were calculated based on the 99th percentile method and also the condition of precipitation coverage. The condition of precipitation coverage in this study is to have precipitation within the 99th percentile and the simultaneous occurrence of heavy precipitation in at least 30% of the stations in the study area (7 stations). Then, the accuracy of 3 precipitation products of GPM-IMERG, TRMM, and AQUA-AIRS satellites was calculated by comparing the heavy rainfall estimated by the satellites with the heavy rainfall recorded by the ground synoptic stations on the days of the event using the Taylor diagram. According to the variable values of the correlation coefficient, standard deviation, and the square root of mean square error in 23 stations according to extreme precipitation events with two temporal and spatial approaches, the precipitation output of each of the satellites was analyzed.



Results and Discussion

According to the calculations made in the study area in the period of 20 years and based on the two conditions considered for the maximum precipitation in this research, 27 maximum precipitation events were identified. GPM-IMERG and AQUA-AIRS satellites performed well in estimating the heavy rainfall in the northwestern region of Iran on a daily time scale and the GPM-IMERG product was more accurate than the other two products. The results also showed that the TRMM product had a greater error in estimating heavy rainfall in the northwestern region of Iran.



Conclusion

The results showed that the GPM-IMERG satellite product can detect heavy rainfall in the northwestern region of Iran and is a good alternative to the rainfall data of ground stations in the region. The results also showed that heavy rainfall estimations by the studied satellites in the west of the research area (West Azarbaijan province) were made with high accuracy.