Accuracy Evaluation And Drought Monitoring Applications Of Precipitation And Evaporation Products In Inner Mongolia Autonomous Region

Precipitation and evaporation are key processes in the global hydrological cycle,and are direct factors leading to natural disasters such as droughts and floods.The acquisition of high spatial and temporal resolution precipitation and evaporation datasets is essential for accurate monitoring and identification of regional droughts.This study evaluated the accuracy of the Tropical Rainfall Measuring Mission(TRMM)Multi-satellite Precipitation Analysis(TMPA)and the Integrated Multi-satellite Retrievals(IMERG)for Global Precipitation Measurement(GPM)dataset,the Global Land Evaporation Amsterdam Model(GLEAM)V3.5a potential evapotranspiration(PET)products,and the Hourly Potential Evapotranspiration Dataset(h PET)based on precipitation,barometric pressure,temperature,sunshine and wind speed and direction of meteorological data from meteorological stations in Inner Mongolia Autonomous Region for the past 20 years,using Correlation Coefficient(CC),Relative Bias(RB),Root Mean Square Error(RMSE),Probability of Detection(POD),False Alarm Rate(FAR)and Critical Success Index(CSI)as accuracy evaluation methods.The precipitation and PET products with the highest accuracy were selected and their performance in drought monitoring applications was explored,with the following main findings:(1)On the daily scale,the correlation of IMERG(CC = 0.71)and the accuracy of precipitation capture(POD=0.76)improved by 16% and 21%,respectively,over TMPA(CC=0.55;POD=0.55),and the algorithmic improvement of IMERG(mean RB= 9%)also significantly mitigates the overestimation of precipitation over complex terrain by TMPA(mean RB = 14%).On the monthly and annual scales,IMERG's CC(monthly: 0.94;annual: 0.92)improves the TMPA(monthly: 0.91;annual: 0.84)by 3%and 8%,respectively,while the RMSE decreases by 2.5 and 30.5,respectively.The correlation of satellite precipitation is lowest in high-altitude desert areas under the influence of complex topographic landscapes,with serious mismeasurement and overestimation of precipitation events.IMERG improves TMPA's severe overestimation of snowfall in winter,but still has the worst correlation.Overall,IMERG outperforms TMPA in detecting precipitation and monsoonal precipitation over complex topographic landscapes,and significantly improves the accuracy of precipitation at different intensities.(2)In the accuracy assessment of annual-scale PET,the CC of h PET(CC=0.81)was much higher than that of GLEAM(CC=0.31),and the RB of h PET from the reference PET(GLEAM:-43.60%;h PET: 6.09%)and the RMSE(GLEAM: 423.4;h PET: 125.8)were smaller.On the monthly and daily scales,the CC of GLEAM(monthly: 0.90;daily: 0.89)and h PET(monthly: 0.95;daily: 0.95)improved over the annual scale,and h PET(monthly RB: 6.09%;daily RB: 7.20%)showed a higher agreement with the reference PET(GLEAM: monthly RB:-45.4%;daily RB:-45.4%).Meanwhile,the daily and monthly products of h PET,have higher precision in the seasonal analysis,with the highest CC(0.92)in spring.In contrast,GLEAM was less consistent in the seasonal analysis,with the lowest CC(0.48)in summer.The GLEAM based on the Priestley-Taylor algorithm is influenced by the type of ground cover and generally underestimates the PET of high-latitude desert regions,and cannot accurately characterize their spatial distribution.In contrast,the h PET based on the PenmanMonteith algorithm can better estimate the PET in the western desert region of Inner Mongolia and well simulate the spatial variation of the reference PET.Overall,the potential evapotranspiration product h PET with high spatial and temporal resolution can accurately characterize the spatial and temporal variation of evapotranspiration in the Inner Mongolia autonomous region.(3)The accuracy of Standardized Precipitation Evapotranspiration Index(SPEI)estimated based on IMERG-F and h PET decreases as the computational time scale increases,and the highest SPEI consistency is found for the one-month time scale(CC=0.87).Spatially,the POD of drought events gradually decreases from east to west,and performs significantly better in Hulunbeier City than in Alashan League and Bayannur City.SPEI is able to capture well the spatial characteristics and temporal variations of four typical drought events(Summer drought event in eastern Inner Mongolia in 2007,spring drought event in northeastern Inner Mongolia in 2009,summer drought event in central and western Inner Mongolia in 2011 and spring and summer consecutive drought events in eastern Inner Mongolia in 2017).The drought events and combined with the drought trend analysis show that Inner Mongolia has a higher frequency of drought events in spring and summer,and the eastern region of Inner Mongolia is prone to severe drought events in summer.However,there is a significant trend of drought aggravation in spring,and there is an overall significant trend of drought weakening in the northeastern region of Inner Mongolia,with most regions in a normal fluctuation.