Assessing the Spatiotemporal Applicability of IsoGSM2 for Simulating Precipitation Isotope Compositions: A Multi-Timescale Analysis

Abstract

Stable isotope in precipitation plays a crucial role in comprehending the water cycle. Isotope-enabled General Circulation Models (iGCMs) can continuously simulate the compositions of stable hydrogen and oxygen isotopes (δ²H and δ¹⁸O) in precipitation. The Isotope-incorporated Global Spectral Model Version 2 (isoGSM2), one of the iGCMs, has been widely used in related research. Here, we use the measured δ¹⁸O records (δ¹⁸O_O) of 21 long-term monitoring stations from the Global Network of Isotopes in Precipitation and modelled δ¹⁸O (δ¹⁸O_M) from isoGSM2 to compare δ¹⁸O_M with δ¹⁸O_O at different time intervals (monthly, annual) for the whole year (MD, AD), rainy season (RSM, RSA) and dry season (DSM, DSA). The results showed that isoGSM2 has good performance in the northern part of Oceania, southeastern Asia, central Europe, eastern North America, Greenland and northern South America across all temporal scales, but the correlations at the other stations vary depending on the timescale. The RMSE and correlation between δ¹⁸O_M and δ¹⁸O_O were better for the rainy season than for the dry season. The relationship between δ¹⁸O_M and temperature was consistent with δ¹⁸O_O for both the MD and RSM, as well as the relationship between δ¹⁸O_M and precipitation amount for the MD. The deviation between simulated and observed values is associated with the accuracy of the model's simulation of rainfall and temperature, as well as the real environmental conditions. In addition, isoGSM2 may have a high correlation at sites where AO and ENSO are significantly affected. These findings are valuable for understanding isoGSM2's applicability and enhance the understanding and knowledge of stable isotopes in the water cycle.