Effects Of Remote Sensing Precipitation Products And Potential Evapotranspiration On Runoff Simulation In Cold Region

Floods,droughts and environmental pollution that have been affected by precipitation all over the world have often caused a great deal of property damage and casualties.As the most important meteorological element,precipitation is a necessary data for climate analysis,drought assessment,flood forecasting,and other related researches.At present,precipitation is mainly observed by ground rainfall stations,weather radar,and remote sensing.Ground rainfall station observation is the most commonly used method for measuring rainfall,but site precipitation is often underrepresented due to the influence of the density of the station network and its spatial distribution.Weather radar has limited observation scope and low coverage,lacks universal applicability and can not meet the needs of rainfall observation in large basins.It has relatively large uncertainties in complex terrain areas.In recent years,the emergence of a series of remote sensing precipitation products with high spatial-temporal resolution has provided new approaches and methods for solving the problems of routine and radar precipitation measurement.Among the representative products,there are mainly TRMM?Tropical rainfall measuring mission?,CMORPH?Climate prediction center morphing technique?,CHIRPS?Climate hazards group infrared precipitation with station data?,GSMaP?Global satellite mapping of precipitation?,etc.Before applying these precipitation products,it is necessary to understand the spatial and temporal distribution of their data errors and their application capabilities in hydrological simulation.At the basin or irrigation area scale,scholars at home and abroad have applied SWAT?Soil and water assessment tool?to carry out more research on hydrological process simulation,non-point source pollution,water resources management and evaluation,and changes in land use methods.However,there are few studies on the effect of potential evapotranspiration?PET?calculated by different methods on the SWAT simulation results,especially the research on the typical watersheds in the high-cold areas like Heilongjiang Province.Heilongjiang Province is the highest latitude and coldest region in China.Winter is cold and long.Snowfall,snowfall,snowmelt,frozen soil,ice-freezing in rivers,river-water flow and other hydrological phenomena are common.These unfavorable factors have increased the research and development difficulty of hydrological models.Up to now,although there are more PET calculation methods,the PET values calculated by different methods are quite different,and each has its own limitations in practical applications.In order to solve the above problems,this paper takes the Hulan River Basin of Heilongjiang Province as the research area,selects TRMM 3B42V7,TRMM 3B42RT,and CHIRPS three types of satellite remote sensing precipitation products,and raster precipitation data are interpolated using by IDW?Inverse distance weighted?interpolation method based on ground rainfall stations.Studying accuracy and error distribution at different time scales and spatial scales of 0.05°×0.05°pixels,and using these precipitation products to drive SWAT to study its applicability in cold runoff simulations,in order to provide more reliable precipitation data for the application of remote sensing precipitation products in cold regions.Taking the Ougen River Basin as a research area,taking into account the natural and geographical conditions such as snowfall and snowmelt hydrological processes and spatial inhomogeneities in the underlying surface,studying on the effect of PET calculated by different methods on the simulation of SWAT daily runoff process in the alpine cold area,in order to find a suitable method for the calculation of the potential evapotranspiration in China's alpine cold regions,and to provide a reference for the application of SWAT in China's alpine cold regions.The results show that 3B42V7,3B42RT,and CHIRPS overestimate 5.43%,41.24%,and 3.37%of the surface precipitation respectively in the watershed daily average scale.3B42V7 is very close to precipitation of the surface observation on the daily,monthly,and seasonal scales of the river basin.The cumulative probability distribution of daily precipitation intensity for 3B42RT is close to rainfall stations.3B42V7 and CHIRPS have better daily runoff simulation during the calibration and verification periods,NSCE was 0.77 and 0.84 in the calibration period,and NSCE was 0.67and 0.56 in the verification period.3B42RT has a poor simulation effect on daily runoff.TRMM3B42V7 and CHIRPS can be considered as basic rainfall data for the study of water resources management in cold regions,drought monitoring,and flood forecasting;SWAT is a coarse way to interpolate PET calculation results on a subbasin scale,which is not a good way to reflect In the spatial distribution of PET,In different subbasins covered by large-scale forests,the monthly average PET value varies slightly with the species and distribution of forest species in different subbasins,it is not that the SWAT's daily runoff results are all good considering the influence of terrain.The Penman-Monteith model Without considering the influence of terrain and the Hargreaces model,the Shuttleworth-Wallace model considering the influence of terrain have better simulation results.The E_ns values of the calibration period and verification period are 0.651 and0.686 respectively.The Penman-Monteith model is more suitable for the calculation of potential evapotranspiration in alpine forest areas.