Statistical evaluation of theoretical and nonparametric approaches to Standardized Precipitation Index

Abu Arra A., Şişman E.

9th Advanced Engineering Days , Tabriz, İran, 9 - 10 Temmuz 2024, ss.1-3

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Basıldığı Şehir: Tabriz
  • Basıldığı Ülke: İran
  • Sayfa Sayıları: ss.1-3
  • Yıldız Teknik Üniversitesi Adresli: Evet

Özet

Accurate drought evaluation is critical for effective water resource management and mitigation strategies. This research aims to compare the nonparametric precipitation drought index using the Standardized Drought Analysis Toolbox (SDAT) tool with the conventional Standardized Precipitation Index (SPI) using different probability distribution functions (CDFs). The comparison employs various statistical metrics, including coefficient of determination (R²), correlation coefficient (CC), mean squared error (MSE), root mean squared error (RMSE), mean absolute error (MAE) and mean bias error (MBE). The Oxford station, with monthly precipitation data spanning from 1767 to 2023, was selected to calculate SPI1 and SPI12. The results highlight the importance of selecting an accurate function for drought analysis. Both theoretical and empirical distributions yielded very similar outcomes; for example, the GEV distribution almost perfectly fit the empirical data (R² = 0.999, CC = 0.999). However, using the SDAT tool offers the advantage of achieving results without error, contributing significantly to the academic field and water resource management.