Cold region SWAT model construction method considering glacier melting process

Abstract

The invention discloses a cold region SWAT model construction method considering a glacier melting process. A calculation method comprises the following steps: step 1, obtaining corrected rainfall data; corrected air temperature data are obtained; evapotranspiration correction data is obtained; actually measured runoff is obtained; 2, obtaining a plurality of sub-basin boundaries and spatial distribution of each sub-basin glacier; 3, calculating the glacier melting amount of each sub-drainage basin in the research time period by adopting a day-to-day factor method to obtain the day-by-day glacier melting amount of each sub-drainage basin; step 4, performing double-element parameter calibration based on the actually measured runoff data and the evapotranspiration correction data to obtain parameters suitable for the watershed hydrological process; and 5, calculating to obtain the relative contribution rate of the glacier melting process to runoff and evapotranspiration. The simulation precision of hydrological elements such as runoff of the cold region hydrological model is improved, and a scientific basis can be provided for sustainable utilization and planning of water resources of a local drainage basin.

Description

Technical Field

[0001] The present invention relates to the technical fields of remote sensing hydrology and cold region hydrology, and specifically relates to a method for constructing a cold region SWAT model considering the glacier melting process. Background Technique

[0002] At present, the calculation methods of glacier runoff are mainly divided into the following five types: direct observation method, glacier mass balance method, water balance method, water chemical tracer method, and hydrological model method, etc. Among them, the hydrological model method is further divided into temperature index model and mass energy balance model. For example, Zhang et al. explored the runoff changes in the Kaidu River Basin under different scenarios based on CMIP6 model data and the VIC (variable infiltration capacity)-glacier model coupled with a glacier module. Gu Huanghe et al. obtained that the contribution rate of glacier meltwater in the Lhasa River to runoff was 14.5% - 21.4% based on the VIC-glacier model, and the contribution rate showed a gradually decreasing trend from upstream to downstream. Jia et al. simulated the glacier hydrological process in the Urumqi No. 1 Glacier Basin based on the glacier hydrological model GHDM (glacio-hydrologic degree-day model), and the results showed that about 51% of the glacier-covered area in this region, and the glacier runoff accounted for about 76% of the total runoff. Wu et al. obtained that the glacier area in the upper reaches of the Shule River decreased by about 21.8% during 1971 - 2020 based on the SPHY (spatial processes in hydrology) model, and the contribution rate of glacier runoff to the total runoff was about 28%.

[0003] The SWAT (soil and water assessment tool) model is the most widely used basin hydrological model. The model improves computational efficiency by dividing sub-basins and hydrological response units (HRUs), and can effectively simulate complex hydrological processes in the basin. The SWAT model does not take into account glacial processes. Adnan et al. believe that the SWAT model does not simulate glacial basins well and should be further combined with glacier modules to improve simulation accuracy. Wei Xiaona et al. [9] embedded the glacier melting algorithm into the SWAT model to simulate the glacier runoff in the upper reaches of the Hotan River Basin, and found that the glacier runoff contribution rates of its two tributaries were 48.7% and 45.5%, respectively. Yang et al. combined the glacier melting algorithm with the SWAT+ model and applied it to the Shache River, a tributary of the Yarkand River. The study showed that glacier runoff is mainly concentrated in summer, accounting for about 78.5% of the annual glacier runoff, and glacier runoff accounts for about 52.5% of the total runoff. The glacier melting algorithms studied above all use the degree-day factor model. After embedding the algorithm, the temperature in the glacier area uses the temperature data of the SWAT model sub-basin. However, glaciers are mostly distributed on the tops of high-altitude mountainous areas and windward slopes, where the temperature is lower than the sub-basin temperature data. Glaciers are very sensitive to temperature response. If this is directly used, it may lead to overestimation of glacier melting and errors in the calculation of glacier melting-related parameters. Therefore, it is necessary to further improve the coupling between the SWAT model and the glacier melting process to improve the applicability of the SWAT model in glacier basins. In addition, most studies focus on the contribution and impact of glacier melting on runoff, ignoring its relationship with basin evapotranspiration.

[0004] Due to the large glacier reserves and abundant observation data, the research on glacier melting in my country is mainly concentrated in the Qinghai-Tibet Plateau, Tianshan Mountains, the northern foot of Kunlun Mountains, Shule River and Heihe River Basins, etc. There are few systematic studies on glacier melting in the Shiyang River Basin. The Shiyang River Basin is located in the eastern section of the Qilian Mountains. The glaciers are small in scale and small in number. They are very sensitive to climate change. It has now reached the "glacier melting inflection point" and is one of the fastest glacier degenerating basins in my country. In addition, the Shiyang River Basin is also one of the areas with the highest population density, high per capita GDP and the most prominent contradiction between people and water in the arid northwest region of my country. Therefore, it is necessary to deeply explore the impact of glacier melting in the Shiyang River Basin on the water cycle process.

[0005] In view of the above problems, the present invention takes the Shiyang River Basin as an example, first corrects the deviation of meteorological data such as precipitation, temperature, and evapotranspiration, drives the glacier melting process based on the corrected temperature data, and then embeds it into the SWAT model which uses the dual elements of runoff and evapotranspiration to simulate and analyze the spatiotemporal evolution of glacier runoff and its impact on runoff and evapotranspiration. It is expected that the present invention can provide a scientific basis for the sustainable utilization and planning of water resources in cold basins covered by glaciers. Summary of the invention

[0006] In order to solve the technical problem described in the background technology, the present invention proposes a cold region SWAT model construction method considering the glacier melting process, comprising the following steps:

[0007] Step 1: bias-correct the obtained precipitation to obtain the corrected precipitation data; bias-correct the obtained temperature to obtain the corrected temperature data; bias-correct the obtained evapotranspiration data to obtain the evapotranspiration corrected data; obtain the measured runoff;

[0008] Step 2: Use ArcGIS software to divide the study area into sub-basins and extract and calculate the glacier area to obtain the boundaries of multiple sub-basins and the spatial distribution of glaciers in each sub-basin;

[0009] Step 3: Based on the spatial distribution of the glaciers obtained in step 2 and the corrected temperature data in step 1, the degree-day factor method is used to calculate the glacier melt amount of each sub-basin within the study period to obtain the daily glacier melt amount of each sub-basin;

[0010] Step 4: Input the daily glacier melt volume obtained in step 3 into the SWAT-glacier model to obtain a processing result; then, process the processing result using SWAT-CUP software, perform dual-factor parameter calibration based on the measured runoff data and the evapotranspiration correction data, and obtain parameters suitable for the hydrological process of the basin;

[0011] Step 5: Substitute the parameters obtained in step 4 into the SWAT-glacier model and the SWAT model, compare the output results of the SWAT-glacier model and the SWAT model, and calculate the relative contribution rate of the glacier melting process to runoff and evapotranspiration.

[0012] Furthermore, the obtained precipitation is subjected to deviation correction to obtain the corrected precipitation data; the obtained temperature is subjected to deviation correction to obtain the corrected temperature data, including:

[0013] Based on the precipitation, the Delta method is used to perform deviation correction on the precipitation to obtain corrected precipitation data; specifically, the remote sensing daily precipitation data is corrected for deviation based on the monthly measured rainfall station data; the formula is as follows:

[0014] P d,i,j＝ P c,i,j ×P a,j / P c,j

[0015] Where P d,i,j is the bias-corrected precipitation data for the jth day of the ith month, P c,i,j is the remote sensing precipitation value of the jth day of the i-th month, P a,j , Pc,j They are the means of measured and remotely sensed precipitation in the j-th month respectively.

[0016] Based on the temperature, the temperature is corrected according to the influence of the altitude factor to obtain corrected temperature data;

[0017] Altitude is an important factor affecting temperature. Since the measured temperature data of the study area is not available, the grid points of CN05.1 temperature data are assumed to be virtual weather stations here. Considering that the temperature decreases with the increase of altitude, first, combined with the elevation data of the virtual weather stations, the temperature data of the weather stations are corrected to sea level height to obtain corrected temperature data. The expression formula is as follows:

[0018] T h = T0 + A × H

[0019] In the formula, T h is the corrected temperature data, in °C; H is the altitude of the virtual weather station, in m; A is the temperature lapse rate, in °C / 100m.

[0020] Based on the corrected temperature data, bilinear interpolation is used to interpolate the corrected temperature data to generate raster data;

[0021] Then, based on bilinear interpolation, the raster data generated by interpolating the corrected temperature data is combined with DEM for correction. The formula is as follows:

[0022] T dem = T s - A × H dem

[0023] In the formula, T dem is the temperature simulation result with a resolution of 1 km after being corrected by DEM, T s is the temperature interpolation result generated by bilinear interpolation, and H dem is the DEM raster data with a resolution of 1 km.

[0024] Furthermore, the deviation correction of the obtained evapotranspiration data to obtain evapotranspiration correction data includes:

[0025] The monthly-scale water balance method of the basin is used to obtain evapotranspiration. The evapotranspiration calculation formula based on the water balance method is as follows:

[0026] ET a = P - R - ΔTWS

[0027] In the formula, ET a is the evapotranspiration calculated based on the water balance method; P is the precipitation, R is the total runoff of the study area, and ΔTWS is the change in terrestrial water storage, which is retrieved by the GRACE gravity satellite. The above units are all in mm.

[0028] Based on the evapotranspiration ET a Evaluate the accuracy of evapotranspiration data and select a Evaporation data with a deviation within a preset deviation threshold; based on the evaporation data with a deviation within the preset deviation threshold, the evaporation data with a deviation within the preset deviation threshold is corrected by using a Delta method to obtain evaporation correction data.

[0029] Furthermore, the daily glacier melt is expressed as the following formula:

[0030] M=PDD*DDF ice

[0031] In the above formula, M is the amount of glacier melt (mm); PDD is the positive accumulated temperature during the study period (℃), calculated from the daily average temperature; DDF ice is the ice degree day factor (mm d -1 ℃ -1 ), can be determined by referring to relevant literature. If there is no reference, the formula can be used for calculation:

[0032] DDF ice =0.009×DEM-0.934×LAT-8.1

[0033] Where DEM is the altitude and LAT is the latitude.

[0034] Furthermore, the step 4 comprises:

[0035] The SWAT model needs to embed the glacier melting process and perform dual parameter calibration of runoff and evapotranspiration to ensure that it is applicable to the hydrological process of the study area. Since the SWAT model does not consider the glacier melting process, the present invention embeds the daily glacier melting amount obtained in step 3 into the model. The idea of ​​embedding is as follows:

[0036] Based on the daily glacier melt, the amount of “precipitation” in the meteorological data is input into the SWAT model, and then the SWAT model is run to obtain the model results;

[0037] Based on the model results, the SUFI-2 program in the SWAT-CUP software was used to perform parameter sensitivity analysis, calibration and verification of the SWAT model. The measured runoff data and remote sensing evapotranspiration data of each sub-basin were used to perform dual-factor and multi-objective parameter calibration. The maximum Nash efficiency coefficient NSE was used as the objective function to obtain the SWAT model parameters suitable for the basin, and the output was the processing result.

[0038] For a river basin, both precipitation and glacier melt are considered as the input components of water resources. The daily-scale glacier melt calculated is added to the daily-scale precipitation and used as the "precipitation" input of the meteorological driving data for the SWAT model. To ensure the matching of the meteorological data and glacier melt input into the SWAT model in terms of format and scale, the central points of each sub-basin in the SWAT model are used as virtual meteorological stations, and the meteorological data corresponding to the virtual meteorological stations of each sub-basin are calculated based on the inverse distance weighting method. Finally, the daily glacier melt and precipitation of the sub-basin are added together and used as the precipitation data of the model to be input into the SWAT model.

[0039] Furthermore, in the two-factor multi-objective calibration of runoff and evapotranspiration, the aggregated NSE value of the SWAT model is calculated according to the following formula:

[0040]

[0041] In the formula, NSE’ is the aggregated NSE value of the SWAT model, and w f is the runoff weight, set to 0.5; NSE f is the simulated NSE value of runoff; n e is the number of target variables of evapotranspiration, and w ej is the evapotranspiration weight of the j-th sub-basin, where the evapotranspiration weights of each sub-basin are equal and the sum is 0.5; NSE ej is the simulated NSE value of evapotranspiration of the j-th sub-basin.

[0042] Furthermore, the relative contribution rates of the glacier melt process to runoff and evapotranspiration satisfy the following formula:

[0043] RC = (x gl - x or ) / x gl

[0044] In the formula, RC is the relative contribution rate, x gl is the runoff / evapotranspiration amount obtained from the SWAT model considering the glacier melt process, and x or is the runoff / evapotranspiration amount obtained from the conventional SWAT model.

[0045] Compared with the prior art, the beneficial effects of the present invention are as follows: The conventional SWAT hydrological model does not consider the influence of the glacier melt process on the basin hydrological process and has poor simulation accuracy in basins with glacier coverage; the present invention combines the SWAT model and the glacier melt process, which can effectively improve the construction accuracy of the hydrological model in cold regions with glaciers and provide a scientific basis for the sustainable utilization and planning of local basin water resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1It is a general overview diagram of the research example of the present invention;

[0047] Figure 2 It is a time series diagram of different evapotranspiration data of the present invention, used to preferably correct evapotranspiration data;

[0048] Figure 3 It is a comparison diagram of the runoff simulation value and the actual value of the model proposed by the present invention;

[0049] Figure 4 It is the relative contribution rate of glacier melting in different basins of the research example to runoff and evapotranspiration. Specific implementation manners

[0050] Next, the technical solution of the present invention in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings of the present invention in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0051] Embodiment:

[0052] A specific use case of the method for constructing a cold region SWAT model considering the glacier melting process of the present invention is as follows: First, it is necessary to perform bias correction on meteorological data. Currently, there are many meteorological data products, but whether they are applicable to the study area needs further verification, and data with strong applicability is preferably selected and corrected. Due to factors such as the error of the sensor itself and the uncertainty of the inversion algorithm, remote sensing products (such as precipitation, evapotranspiration, etc.) have certain biases in mountainous areas with complex terrain; the number of measured stations in the study area is scarce, and the CN05.1 temperature data interpolated based on measured data does not consider the influence of DEM on temperature. However, for a hydrological model, the accuracy of meteorological data is the key to improving the model simulation accuracy. Therefore, before running the SWAT model, it is necessary to perform bias correction on precipitation, temperature, and evapotranspiration data.

[0053] Precipitation is the main source of water resources in the basin. Here, the measured data of monthly rainfall stations are used to perform bias correction on the daily-scale precipitation data of Chirsp. The formula is as follows:

[0054] P d,i,j＝ P c,i,j ×P a,j / P c,j

[0055] In the formula, P d,i,j is the precipitation data of the jth day of the ith month after bias correction, P c,i,j is the remote sensing precipitation value of the jth day of the ith month, P a,j 、P c,jare the means of measured and remotely sensed precipitation in the jth month, respectively.

[0056] Altitude is an important factor affecting temperature. Since we do not have the actual temperature data in the study area, we assume that each grid point of CN05.1 is a virtual weather station. Considering that the temperature decreases with increasing altitude, we first correct the temperature data of the weather station to the sea level by combining the elevation data of the virtual weather station. The formula is as follows:

[0057] T h =T0+A×H

[0058] Where, T h is the temperature corrected to sea level, ℃; H is the altitude of the virtual weather station, m; A is the temperature lapse rate, ℃ / 100m.

[0059] Then, the corrected data is interpolated based on bilinear interpolation to generate raster data combined with DEM correction. The formula is as follows:

[0060] T dem =T s -A×H dem

[0061] Where, T dem is the temperature simulation result with a resolution of 1 km after DEM correction, T s is the temperature interpolation result generated by bilinear interpolation, H dem The DEM raster data is 1km resolution.

[0062] The adaptability of remote sensing evapotranspiration data in the study area varies greatly, so accuracy comparison and bias correction are required before application. a The accuracy of remote sensing evapotranspiration data was evaluated by using the ET a The evapotranspiration data with deviations within the preset deviation threshold are then corrected based on the Delta method. The actual evapotranspiration calculation formula based on the water balance method is as follows:

[0063] ET a =PR-ΔTWS

[0064] In the formula, ET a is the evapotranspiration calculated based on the water balance method; P is the precipitation, R is the total runoff of the study area, and ΔTWS is the change in terrestrial water storage, which is inverted by the GRACE gravity satellite. All the above units are mm.

[0065] Then, the amount of glacier melt is calculated. The degree-day factor model is used to calculate the amount of glacier melt, and the formula is as follows:

[0066] M=PDD*DDF ice

[0067] Where M is the amount of glacier melt (mm); PDD is the positive accumulated temperature during the study period (℃), calculated from the daily average temperature; DDF ice is the ice degree day factor, which can be determined by referring to relevant literature. If it is not available, it can be calculated by referring to the formula:

[0068] DDF ice =0.009×DEM-0.934×LAT-8.1

[0069] Where DEM is the altitude and LAT is the latitude.

[0070] The sub-basin division was completed based on the SWAT model, and the glacier area extraction was performed based on ArcGIS.

[0071] In the construction of the SWAT model, since the model does not consider the glacier melting process, the present invention aggregates the daily glacier melting amount calculated above to each sub-basin to obtain the daily glacier melting amount of each sub-basin. Subsequently, in order to ensure that the meteorological data input into the SWAT model and the glacier melting amount match in format and scale, the center point of each sub-basin is used as a virtual meteorological station in the SWAT model, and the meteorological data of the virtual meteorological station corresponding to each sub-basin is calculated based on the inverse distance weighted method; finally, the daily glacier melting amount of the sub-basin is added to the precipitation, and input into the SWAT model as the precipitation data of the model.

[0072] In the model parameter calibration part, the mountain runoff and the actual evapotranspiration of each sub-basin were selected for multi-objective calibration. The SUFI-2 program in the SWAT-CUP software was used for parameter sensitivity analysis, calibration and verification, and the t-Stat value was used to evaluate the parameter sensitivity. The larger the absolute value of the t-test, the more sensitive the parameter. When calibrating the parameters, the maximum Nash efficiency coefficient NSE was selected as the objective function. In the dual-factor multi-objective calibration of runoff and evapotranspiration, the NSE summary value of the SWAT model was calculated according to the following formula:

[0073]

[0074] Where NSE' is the NSE summary value of the SWAT model, w f is the runoff weight, which is set to 0.5 according to previous studies; NSE f is the simulated NSE value of runoff; n e is the number of target variables for evapotranspiration, w ej is the evapotranspiration weight of the jth sub-basin, where the evapotranspiration weights of each sub-basin are equal and the sum is 0.5; NSE ej is the simulated NSE value of evapotranspiration of the jth sub-basin.

[0075] The relative contribution rates of the glacier melting process to runoff and evapotranspiration are as follows:

[0076] RC = (x gl - x or ) / x gl

[0077] In the formula, RC is the relative contribution rate, and x gl is the runoff / evapotranspiration amount obtained from the SWAT model considering the glacier melting process, and x or is the runoff / evapotranspiration amount obtained from the conventional SWAT model. Specific embodiments:

[0079] (1) Study area

[0080] The Shiyang River Basin (36°29’ - 39°27’N, 101°41’ - 104°16’E) is located in the eastern part of the Hexi Corridor in Gansu Province and is one of the three inland river basins in the Hexi Corridor. Affected by the terrain obstruction, the Qilian Mountains in the upper reaches of the basin have abundant precipitation and glacier development. It is the origin of the Shiyang River and an important water conservation area with a high vegetation coverage rate. The precipitation in the middle and lower reaches is scarce and the evaporation is strong, belonging to a typical resource-based water shortage area. Affected by climate change and human activities, the water resources in the basin are increasingly scarce, the ecological water use is difficult to guarantee, and it faces a serious ecological crisis. There are a total of 8 tributaries in the upper reaches of the Shiyang River Basin, and the tributaries with glaciers from west to east are the Dongda River, Xiying River, Jinta River, and Zamu River respectively. Figure 1 It is a general situation map of the Shiyang River Basin. The observation time of the glacier boundary in the figure is 1980. It can be seen from the figure that the glacier scale in the Dongda River Basin is relatively large, and the glacier scales in the Jinta River and Zamu River basins are relatively small.

[0081] (2) Changes in glacier area and scale

[0082] Table 1 shows the changes in the glacier area of the Shiyang River Basin. It can be seen that from 1980 to 2015, the glacier area in the Shiyang River Basin showed a significant decreasing trend, from 60.10 km 2 sharply reduced to 33.61 km 2 in 2015. Among them, the glacier area of the Dongda River in the basin is the largest, accounting for about 50% of the Shiyang River Basin, followed by the Xiying River, Nanying River, and Zamu River. Table 2 shows the changes in the glacier scale of the Shiyang River Basin. It can be seen that the number of glaciers with an area ≤ 1 km 2 dominates in the Shiyang River Basin. Glaciers of this scale are relatively sensitive to climate change due to their relatively poor stability. Affected by climate change, the glacier ablation phenomenon is obvious. From 1980 to 2015, the number of glaciers with an area ≤ 0.1 km 2 increased by 28, and the area increased by 1.06 km2 , while the number of glaciers > 0.5 km 2 decreased by 23, and the area decreased by 25.14 km 2 .

[0083] Table 1 Glacier area changes in the Shiyang River Basin from 1980 to 2015 Unit: km 2

[0084]

[0085] Table 2 Glacier size changes in the Shiyang River Basin from 1980 to 2015

[0086]

[0087] (2) Screening and calibration of evapotranspiration data

[0088] In this case, a variety of remote sensing evapotranspiration data were used, such as GLASS (ET GLASS ), MODIS (ET MODIS ), ETmonitor (ET monitor ), GLDAS (ET GLDAS ), PML-ET (ET PML ), etc. Subsequently, screening and calibration were carried out based on the actual evapotranspiration (ET a ) obtained from the water balance equation, and the results are as Figure 2 shown. From 2003 to 2018, the average annual values of ET a , ET PML , ET GLDAS , ET monitor , ET MODIS and ET GLASS were 347.4, 395.8, 453.5, 495.1 and 342.3 mm respectively. In summary, the accuracy of ET GLASS in the upper reaches of the Shiyang River Basin is significantly better than other evapotranspiration data. Therefore, ET GLASS was selected as the calibration data for evapotranspiration in the SWAT model. To further improve the accuracy of model evapotranspiration, the Delta method was used to calibrate the ET GLASS model here, and the calibration coefficient was calculated to be 1.015.

[0089] (3) SWAT model evaluation

[0090] Due to space limitations, only the Dongda River Basin, which has the largest proportion of glacier area in the Shiyang River Basin, will be taken as an example for presentation in the following. Considering the availability of runoff data, the warm-up period of the model is set as 1981 - 1985, the calibration period and validation period of the runoff rate of the Dongda River are 1986 - 2005 and 2006 - 2021 respectively, and the calibration period and validation period of the evapotranspiration rate are from March 2000 to December 2010 and from 2011 to 2018 respectively.

[0091] Based on the SUFI2 algorithm, parameter sensitivity analysis is carried out, and the top 10 most sensitive parameters of the basin are shown in Table 3.

[0092] Table 3 Ranking of parameter sensitivity of the model

[0093]

[0094] The optimal parameters obtained during the calibration period of the SWAT - glacier model are substituted into the validation period and the SWAT model for simulation and validation. The relevant results can be seen in Table 4, where the evaluation index value of evapotranspiration is the weighted average calculated according to the sub - basin area. The results show that compared with the conventional SWAT model, the SWAT - glacier model has a certain improvement in the simulation accuracy of runoff and evapotranspiration. The NSE values of runoff and evapotranspiration of the Dongda River in the validation period have increased by about 9.9% and 3.9% respectively. The NSE, R 2 and PBIAS values of the runoff simulated during the validation period of the SWAT - glacier model are 0.579, 0.619 and 15.236% respectively. The time series of the monthly runoff simulation values and actual values during the calibration period and validation period based on the SWAT - glacier model are as Figure 3 shown.

[0095] Table 4 Evaluation indexes of the model

[0096]

[0097] (4) Impact of glacier melt on runoff and evapotranspiration

[0098] Figure 4 The relative contributions of glacier melt to runoff and evapotranspiration in the Dongda River Basin from 1986 to 2021 are shown. It can be seen that due to the continuous reduction of glacier area, the impact of glacier melt on runoff and evapotranspiration shows a fluctuating decreasing trend. In addition, the peaks and troughs of the contribution rates of runoff and evapotranspiration basically correspond to each other. This is because in years with larger precipitation, the contribution rate of glaciers to runoff is relatively reduced, and abundant water will cause an increase in evapotranspiration; the opposite is true in years with less precipitation. The multi - year average relative contribution rates of glacier melt to runoff and evapotranspiration in the Dongda River Basin are 6.5% and 8.0% respectively.

[0099] The above embodiments are only used to illustrate the present invention and are not intended to limit the present invention. Those of ordinary skill in the relevant technical field can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all equivalent technical solutions also fall within the scope of the present invention. The patent protection scope of the present invention is defined by the claims.

[0100] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for constructing a cold region SWAT model considering the glacier melting process, characterized in that: The steps include: Step 1: bias-correct the obtained precipitation to obtain the corrected precipitation data; bias-correct the obtained temperature to obtain the corrected temperature data; bias-correct the obtained evapotranspiration data to obtain the evapotranspiration corrected data; obtain the measured runoff; Step 2: Use ArcGIS software to divide the study area into sub-basins and extract and calculate the glacier area to obtain the boundaries of multiple sub-basins and the spatial distribution of glaciers in each sub-basin; Step 3: Based on the spatial distribution of the glaciers obtained in step 2 and the corrected temperature data in step 1, the degree-day factor method is used to calculate the glacier melt amount of each sub-basin within the study period to obtain the daily glacier melt amount of each sub-basin; Step 4: Input the daily glacier melt volume obtained in step 3 into the SWAT-glacier model to obtain a processing result; then, process the processing result using SWAT-CUP software, perform dual-factor parameter calibration based on the measured runoff data and the evapotranspiration correction data, and obtain parameters suitable for the hydrological process of the basin; Step 5: Substitute the parameters obtained in step 4 into the SWAT-glacier model and the SWAT model, compare the output results of the SWAT-glacier model and the SWAT model, and calculate the relative contribution rate of the glacier melting process to runoff and evapotranspiration.

2. The method according to claim 1, characterized in that The method includes performing deviation correction on the obtained precipitation to obtain the corrected precipitation data; and performing deviation correction on the obtained temperature to obtain the corrected temperature data, including: Based on the precipitation, a Delta method is used to perform deviation correction on the precipitation to obtain corrected precipitation data; Based on the temperature, the temperature is corrected according to the influence of the altitude factor to obtain corrected temperature data.

3. The method according to claim 1, characterized in that The method of performing deviation correction on the acquired evapotranspiration data to obtain evapotranspiration correction data includes: The monthly water balance method of the basin is used to obtain evapotranspiration. The evapotranspiration calculation formula based on the water balance method is as follows: AND a =PR-ΔTWS In the formula, ET a is the evapotranspiration calculated based on the water balance method; P is the precipitation, R is the total runoff in the study area, and ΔTWS is the change in terrestrial water storage, which is obtained by inversion of the GRACE gravity satellite; Based on the evapotranspiration ET a Evaluate the accuracy of evapotranspiration data and select a Evaporation data with a deviation within a preset deviation threshold; based on the evaporation data with a deviation within the preset deviation threshold, the evaporation data with a deviation within the preset deviation threshold is corrected by using a Delta method to obtain evaporation correction data.

4. The method according to claim 1, characterized in that: The daily glacier melt is expressed as follows: M=PDD*DDF ice In the above formula, M is the amount of glacier melt; PDD is the positive accumulated temperature during the study period; DDF ice It is the ice degree day factor.

5. The method according to claim 1, characterized in that The step 4 comprises: Based on the daily glacier melt, the amount of "precipitation" in the meteorological data is input into the SWAT model, and then the SWAT model is run to obtain the model results; Based on the model results, the SUFI-2 program in the SWAT-CUP software was used to perform parameter sensitivity analysis, calibration and verification of the SWAT model. The measured runoff data and remote sensing evapotranspiration data of each sub-basin were used to perform dual-factor and multi-objective parameter calibration. The maximum Nash efficiency coefficient NSE was used as the objective function to obtain the SWAT model parameters suitable for the basin, and the output was the processing result.

6. The method according to claim 5, characterized in that In the dual-factor multi-objective calibration of runoff and evapotranspiration, the NSE summary value of the SWAT model is calculated according to the following formula: Where NSE' is the NSE summary value of the SWAT model, w f is the runoff weight; NSE f is the simulated NSE value of runoff; n e is the number of target variables for evapotranspiration, w ej is the evapotranspiration weight of the jth sub-basin; NSE ej is the simulated NSE value of evapotranspiration of the jth sub-basin.

7. The method according to claim 1, characterized in that The relative contribution of the glacier melting process to runoff and evapotranspiration satisfies the following formula: RC=(x gl -x or ) / x gl In the formula, RC is the relative contribution rate, x gl is the runoff / evapotranspiration obtained by the SWAT model considering the glacier melting process, x or is the runoff / evapotranspiration obtained by the conventional SWAT model.

Images