ZHANG Le-le, GAO Li-ming, CHEN Ji, ZHAO Lin, CHEN Ke-long, ZHAO Jing-yi, LIU Guo-jun, SONG Ting-xi, LI Yan-kun
2022, 19(07): 1974-1987.
Accurately measuring precipitation is integral for understanding water cycle processes and assessing climate change in the Qinghai–Tibet Plateau(QTP). The Geonor T-200B weighing precipitation gauge with a single Alter shield(Geonor) and the Chinese standard precipitation gauge(CSPG) are widely used for measuring precipitation in the QTP. However, their measurements need to be adjusted for wetting loss, evaporation loss and windinduced undercatch. Four existing transfer functions for adjusting the Geonor-recorded and three transfer functions for adjusting the CSPG-recorded precipitation at hourly, daily or event scale has been proposed based on the precipitation intercomparison experiments conducted at a single site in different regions. Two latest transfer functions for the Geonor (which are referred to as K2017a and K2017b) at the half-hour time scale based on the precipitation intercomparison experiments at multiple stations in the northern hemisphere were provided in the World Meteorological Organization Solid Precipitation Intercomparison Experiment. However, the applicability of these transfer functions in the QTP has not been evaluated. Therefore, the current study carried out a precipitation measurement intercomparison experiment between August 2018 and September 2020 at a site in Beiluhe in central QTP. The performance of these transfer functions at this site was also evaluated on the basis of mean bias(MB), root mean squared error(RMSE) and relative total catch(RTC). The results are as follows: First, the unadjusted RTC values of the Geonor for rain, mixed(snow mixed with rain), snow and hail are 92.06%, 85.35%, 64.11% and 91.82%, respectively, and the unadjusted RTC values of the CSPG for the same precipitation types are 92.59%, 81.32%, 46.43% and 95.56%, respectively. Second, K2017a has the most accurate adjustment results for the Geonor-recorded snow and mixed precipitation at the half-hour time scale, and the post-adjustment RTC values increased to 98.25% and 98.23%, respectively. M2007e, an event-based transfer function, was found to have the most accurate adjustment results for the Geonorrecorded snow precipitation at the event scale, and the post-adjustment RTC value increased to 96.36%. Third, the existing transfer functions for CSPG underestimate snowfall, while overestimating rainfall. Fourth, hail is a significant precipitation type in central QTP. The catch efficiency of hail precipitation and the temperature when hail precipitation occurs are close to those of rain; moreover, the transfer functions for rain are suitable for hail as well.
