|
Evaluation of GPM-IMERG precipitation estimate errors using empirical quantile mapping for runoff management in the Caspian Coastal Region
|
Bromand Salahi *    , Ali Shahi |
| University of Mohaghegh Ardabili |
|
|
Abstract: (508 Views) |
| This study aimed to evaluate and bias-correct GPM-IMERG satellite precipitation estimates using the Empirical Quantile Mapping (EQM) method in the Caspian Sea coastal region of Iran over the 2005–2014 period. To this end, daily precipitation data from five synoptic stations—Bandar Anzali, Rasht, Ramsar, Babolsar, and Gorgan—representing the diverse climatic characteristics, were used. Ground-based observations were obtained from the Iran Meteorological Organization (IRIMO), while GPM-IMERG satellite data were retrieved via the Google Earth Engine (GEE) platform. The performance of the satellite estimates and the effectiveness of the bias correction were evaluated at daily, monthly, and annual timescales using several statistical metrics, including the Pearson correlation coefficient, Root Mean Square Error (RMSE), Nash–Sutcliffe Efficiency (NSE), Percent Bias (PBIAS), Probability of Detection (POD), and False Alarm Ratio (FAR). Additionally, Taylor diagrams were employed to simultaneously assess correlation, standard deviation, and root-mean-square error, analyzing satellite performance before and after bias correction. The results indicated that applying the EQM method improved GPM-IMERG performance at most stations across both daily and monthly timescales. At the daily scale, RMSE decreased by 5.7% to 20.4% across all stations. Furthermore, the absolute percent bias (∣PBIAS∣) converged toward zero, showing reductions ranging from 47.1% to 98.9%. The NSE values also improved across all stations, though they remained negative in most cases at the daily level. On the monthly scale, the impact of the bias correction was more pronounced; RMSE decreased by 6.2% to 27.9% in four out of five stations, although it increased by 4.6% in Bandar Anzali. Additionally, ∣PBIAS∣ decreased across all stations, with reductions ranging from 56.5% to 98.6%. Moreover, monthly NSE exhibited notable improvements, rising from 0.43 to 0.70 in Gorgan, 0.29 to 0.41 in Rasht, and -0.29 to 0.13 in Ramsar. |
Article number: 7 |
| Keywords: Bias Correction, Northern Iran, Remote Sensing, Satellite Data, Statistical Evaluation. |
|
|
|
Type of Study: Applicable |
Subject:
Special
Received: 2026/06/8 | Revised: 2026/09/11 | Accepted: 2026/07/30 | ePublished ahead of print: 2026/07/31 | ePublished: 2026/07/31
|
|
|
|
|
|
|
| Send email to the article author |
|
|
|
| Add your comments about this article |
|
|
|
