A method for simulating and calculating runoff in the source area of ​​the Yellow River

Abstract

The present invention discloses a method for simulating and calculating runoff in the source region of the Yellow River, comprising the following steps: selecting a WRF-Hydro model according to the actual situation of the watershed, collecting model input data and observation data of the watershed, comparing the accuracy of four types of precipitation products and obtaining final driving data of the model, and simulating and outputting runoff calculation results using the WRF-Hydro model; the present invention compares the accuracy of the four types of precipitation products at different temporal and spatial scales, selects the optimal precipitation data and combines it with the non-precipitation field of GLDAS to obtain the final driving data of the model, drives the WRF-Hydro model through the input of each phase data, and then simulates and calculates the runoff of the watershed to be simulated. The method can accurately and effectively simulate the runoff changes in the watershed, reduce the uncertainty of the simulation results, and thus improve the accuracy of the runoff calculation results.

Description

Technical Field

[0001] The present invention relates to the technical field of regional water resources management, and in particular to a method for simulating and calculating runoff in the source area of ​​the Yellow River. Background Art

[0002] The Yellow River source region, located in the heart of the Qinghai-Tibet Plateau, is influenced by both the plateau monsoon and the East Asian monsoon, resulting in a typical plateau subarctic semi-arid continental monsoon climate with extremely uneven precipitation distribution. As the primary runoff-generating and water conservation area in the Yellow River basin, river runoff in the source region is primarily composed of precipitation and glacial meltwater. The multi-year average runoff at Tangnaihai, the controlling hydrological station, is 209.3 × 108 m3, accounting for approximately 34.1% of the Yellow River's total natural runoff. In recent years, global warming has led to rising temperatures, intensified evaporation, and decreased precipitation in the source region. Significant changes have occurred in water conservation units such as glaciers and permafrost, significantly impacting runoff. Therefore, accurately estimating the magnitude of runoff in the Yellow River source region and revealing its evolutionary patterns, establishing a corresponding land-surface hydrological coupling model for runoff simulation, and improving the accuracy of runoff prediction are of great significance for regional water resources management and ecological sustainable development.

[0003] Regional precipitation is the most important meteorological factor affecting runoff. Its spatiotemporal distribution and variations directly determine the degree of dryness and wetness in a region. Accurate precipitation data are essential for reliable runoff simulation. However, due to various limitations, such as climatic conditions, underlying surface characteristics, and human activities, the density of ground rain gauges is low and their distribution is uneven. Missing and mismeasured measurements are common, making it difficult to accurately characterize the true spatial distribution characteristics of precipitation. This poses certain difficulties for basin hydrological simulation and forecasting. Currently, commonly used high-spatiotemporal resolution precipitation products at home and abroad include the China Meteorological Forcing Dataset (CMFD), the Global High-Resolution Precipitation Dataset (CPCMorphing technique, CMORPH), the Tropical Rainfall Measuring Mission (TRMM), and the Global Land Data Assimilation System (GLDAS).

[0004] The Weather Research and Forecasting Model Hydrological modeling system (WRF-Hydro) is a high-resolution distributed land-atmosphere coupled model developed by the National Center for Atmospheric Research in the United States to improve the redistribution of surface, underground, and river water and promote the coupling of atmospheric and hydrological models. The WRF-Hydro model can operate as an independent land surface hydrological model or be coupled with atmospheric models, such as mesoscale weather forecasts, to achieve a two-way feedback process between the atmosphere and the land surface. Compared with traditional land surface hydrological models, the WRF-Hydro model is designed to provide continuous spatial gridded information such as soil temperature and humidity, evapotranspiration flux, water and heat exchange flux, and runoff.

[0005] Most existing runoff simulation and calculation methods have complicated steps, and the precipitation-driven data is not accurate, resulting in the inability to accurately and effectively simulate runoff changes. The simulation results have high uncertainty, resulting in inaccurate runoff calculation results. They cannot provide assistance for regional large-scale basin hydrological simulation, runoff prediction and regional water resources management. Therefore, the present invention proposes a method and system for simulating and calculating runoff in the Yellow River source area to solve the problems existing in the existing technology. Summary of the Invention

[0006] In view of the above problems, the purpose of the present invention is to propose a method and system for simulating and calculating the runoff in the source area of ​​the Yellow River, so as to solve the problem that the existing runoff simulation and calculation methods lead to inaccurate runoff calculation results.

[0007] To achieve the purpose of the present invention, the present invention is implemented by the following technical solution: a method for simulating and calculating the runoff in the source area of ​​the Yellow River, comprising the following steps:

[0008] Step 1: Select the WRF-Hydro model according to the actual situation of the watershed to be simulated. This model includes the land surface module, soil confluence module, surface confluence module, groundwater confluence module, and river confluence module;

[0009] Step 2: First, collect the meteorological driving data, underlying surface data, and river network data of the watershed to be simulated as model input data, and perform bilinear interpolation on the obtained GLDAS data to obtain the initial driving data of the watershed to be simulated;

[0010] Step 3: First, obtain precipitation data for the simulated basin from four precipitation products: CMFD, CMORPH, TRMM, and GLDAS. Then, compare the accuracy of the precipitation data from these four precipitation products in the simulated basin. Select the precipitation data with the best accuracy and combine it with the non-precipitation field from GLDAS to obtain the final driving data for the WRF-Hydro model.

[0011] Step 4: Input the final driving data of the watershed to be simulated and the initial field data of the model operation into the WRF-Hydro model, then run the WRF-Hydro model. The model land surface module calculates the runoff of the input relevant water body data, and then downscales the surface runoff and soil moisture data to the confluence grid through the grid for calculation, and finally obtains the runoff of the watershed to be simulated.

[0012] A further improvement is that in step 1, the soil confluence module is a three-dimensional calculation module that takes into account water exchange in both vertical and horizontal directions. When the WRF-Hydro mode is running, the land surface module on the coarse grid is first calculated, and then the confluence module on the fine grid is calculated, including the soil confluence module, the surface confluence module, the groundwater confluence module and the river confluence module.

[0013] A further improvement is that in step 1, in the WRF-Hydro model, the global reanalysis data FNL is used as the initial field for the WRF-Hydro model operation, and FNL is input into the WRF model to obtain the initial field state for the WRF-Hydro model operation.

[0014] Further improvements are as follows: in step 1, the temporal and spatial resolutions of the WRF-Hydro model driving data are 3 h and 5.0 km × 5.0 km respectively, the confluence grid resolution is 500.0 m × 500.0 m, the calculation time step is set to 20 s, and the runoff simulation results are output hourly.

[0015] A further improvement is that in step 2, the meteorological driving data consists of seven variables: downward longwave radiation, downward shortwave radiation, surface air pressure, specific humidity, air temperature, near-surface wind speed, and precipitation rate. The temporal and spatial resolutions of the GLDAS data are 3 h and 0.25°×0.25°, respectively.

[0016] Further improvements are as follows: in step 2, the underlying surface data are all from the WRF pre-processing system, the river network data are from hydrological data and maps based on multi-scale elevation derivatives, elevation data with a resolution of 90.0m×90.0m are selected, and accurate river network information is extracted.

[0017] A further improvement is that in step three, when performing accuracy comparison, the precipitation at the basin station to be simulated is first interpolated into 0.25°×0.25° grid data using the Cressman objective analysis method, and then the four types of precipitation products CMFD, CMORPH, TRMM and GLDAS are interpolated to the same resolution using bilinear interpolation. Then, the spatial accuracy is compared using the grid data, and the bilinear interpolation method is further used to interpolate the grid data to the station. The arithmetic mean of the station is used to represent the precipitation in the basin to be simulated, and the temporal accuracy of the precipitation products is compared.

[0018] Further improvements are as follows: in step three, four evaluation indicators, namely correlation coefficient, root mean square error, Nash coefficient and relative deviation, are selected to judge the accuracy of precipitation. At the same time, three statistical indicators, namely detection rate, false alarm rate and detection success rate, are used to evaluate the ability of different precipitation products to capture precipitation in the Yellow River source area.

[0019] A further improvement is that in step 4, the final driving data of the watershed to be simulated are precipitation data from CMFD and non-precipitation data from GLDAS, and the initial field data required for the WRF-Hydro model are provided by FNL data.

[0020] The beneficial effects of the present invention are as follows: the present invention compares the precision of precipitation data of four types of precipitation products, namely CMFD, CMORPH, TRMM and GLDAS, and obtains the final driving data of the WRF-Hydro model in combination with the non-precipitation field of GLDAS, and drives the WRF-Hydro model through the input of each phase data, thereby simulating and calculating the runoff of the basin to be simulated. The present invention can accurately and effectively simulate the runoff changes in the basin and reduce the uncertainty of the simulation results, thereby improving the accuracy of the runoff calculation results, and at the same time provides important reference value for large-scale basin hydrological simulation, runoff prediction and regional water resources management. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0022] Figure 1 It is a schematic flow chart of the method of the present invention;

[0023] Figure 2 Schematic diagram of monthly runoff changes in an embodiment of the present invention. DETAILED DESCRIPTION

[0024] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0025] See also Figure 1 、 Figure 2 This embodiment provides a method for simulating and calculating the runoff in the source area of ​​the Yellow River, including the following steps:

[0026] Step 1: Select the Yellow River source area as the basin to be simulated, with a latitude and longitude range of 32.12~35.48°N, 95.50~103.28°E, and an altitude between 2775.0 and 6164.0m. Select the WRF-Hydro5.1.1 version in the uncoupled mode as the operation mode. The temporal and spatial resolutions of the model driving data are 3h and 5.0km×5.0km respectively, the confluence grid resolution is 500.0m×500.0m, the calculation time step is set to 20s, and the runoff simulation results are output hourly. This model includes The WRF-Hydro model includes the land surface module, the subsoil confluence module, the surface confluence module, the groundwater confluence module, and the river confluence module. The subsoil confluence module is a three-dimensional calculation module that considers both vertical and horizontal water exchange. When the WRF-Hydro model is running, the land surface module is calculated first, and then the confluence modules on the fine grid, namely the subsoil confluence module, the surface confluence module, the groundwater confluence module, and the river confluence module, are calculated. Before the WRF-Hydro model is run, the global reanalysis data FNL is input into the WRF model to obtain the initial field state for the WRF-Hydro model.

[0027] Step 2: First, precipitation data from nine meteorological observation stations in the simulated basin from 2009 to 2018 and runoff data from the Tangnaihai hydrological station were collected. Gridded meteorological driving data, underlying surface data, and river network data were used as model input data. The meteorological driving data consisted of seven variables: downward longwave radiation, downward shortwave radiation, surface pressure, specific humidity, air temperature, near-surface wind speed, and precipitation rate. Bilinear interpolation was performed on GLDAS data with a temporal and spatial resolution of 3 hours and 0.25°×0.25° for the simulated basin, respectively, to obtain the initial driving data for the simulated basin. This data was jointly developed by NASA and the National Centers for Environmental Prediction and integrates ground and satellite observations.

[0028] Step 3: First obtain the precipitation data of the four types of precipitation products, CMFD, CMORPH, TRMM and GLDAS, in the basin to be simulated. Then compare the accuracy of the precipitation data of the four types of precipitation products, CMFD, CMORPH, TRMM and GLDAS, in the basin to be simulated. Select the precipitation data with the best accuracy and combine it with the non-precipitation field of GLDAS to obtain the final driving data of the WRF-Hydro model. When comparing the accuracy, first use the Cressman objective analysis method to interpolate the precipitation of the basin station to be simulated into 0.25°×0.25° grid data, and then use bilinear interpolation to interpolate the CMFD, CMORPH, TRMM and GLDAS grid data. The four precipitation products of RMM and GLDAS were interpolated to the same resolution (0.25°). The spatial accuracy was then compared using gridded data. The gridded data were further interpolated to the stations using the bilinear interpolation method. The arithmetic mean of the stations was used to represent the basin precipitation. The temporal accuracy of the precipitation products was compared. Four evaluation indicators, including correlation coefficient (R), root mean square error (RMSE), Nash coefficient (NSE), and relative bias (BIAS), were selected to evaluate the precipitation accuracy. Three statistical indicators, including detection rate (POD), false alarm rate (FAR), and detection success rate (CSI), were used to evaluate the ability of different precipitation products to capture precipitation in the Yellow River source region.

[0029] Step 4: The vegetation type, land use type, and soil type information required by the WRF-Hydro model are all obtained from the WRF preprocessing system. The default soil type is replaced by the more accurate soil type dataset from Beijing Normal University. The high-resolution river network data required by the WRF-Hydro model are obtained from hydrological data and maps based on multi-scale elevation derivatives. Elevation data with a resolution of 90.0 m × 90.0 m are selected, and accurate river network information is extracted. The final driving data for the simulated watershed and the initial field data for the model run are input into the WRF-Hydro model. The WRF-Hydro model is then run to simulate the runoff of the simulated watershed. The model calculates runoff generation from the input relevant water body data through the land surface module. The surface runoff and soil moisture data are then downscaled to the confluence grid for calculation. The final runoff for the simulated watershed is obtained, and the Nash coefficient (NSE), root mean square error (RMSE), and correlation coefficient (R) are used to evaluate the quality of the runoff simulation results.

[0030] The detailed information of each data in this embodiment is shown in Table 1 below

[0031] Table 1 Data Detailed Information Display Table

[0032]

[0033] The precipitation data accuracy and runoff simulation performance evaluation in this embodiment are shown in Table 2 below.

[0034] Table 2 Precipitation data accuracy and runoff simulation performance evaluation table

[0035]

[0036] In Table 2, N represents the number of samples, S i and O i They represent precipitation products and observed precipitation, respectively. A represents the days when both precipitation products and measured data show precipitation, B represents the days when the reprecipitation product shows precipitation but the observation data shows no precipitation, and C represents the days when the observation data shows precipitation but the precipitation product shows no precipitation.

[0037] Based on the WRF-Hydro model simulation and calculation, the monthly runoff changes at the Tangnaihai hydrological station in the Yellow River source area from 2009 to 2018 are as follows: Figure 2 As shown in the data, the simulation results during the calibration period (2012-2013) were good, with correlation coefficients (R) above 0.95 and Nash coefficients (NSE) above 0.92. Overall, the simulated runoff during the test period was consistent with the observed values, and the flow process lines were relatively consistent. However, the flood peak error was large during the flood season, which was related to the relatively large precipitation concentration in the CMFD data area. The wet and dry years in each validation period led to large differences in the simulation results. The simulation effect in wet years was generally stronger than that in dry years. In particular, in 2016, the overall simulation results were poor, with large errors in the runoff process lines and flow size compared with the observations, and the Nash coefficient (NSE) was only 0.15. Therefore, the parameter combination obtained based on calibration in wet years has certain limitations in applicability to dry years. Using higher-resolution underlying surface and driving data and considering multi-parameter calibration methods will further improve the simulation results.

[0038] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for simulating and calculating runoff in the source area of ​​the Yellow River, characterized in that: The following steps are involved: Step 1: Select the WRF-Hydro mode according to the actual situation of the basin to be simulated. The mode includes a land surface module, a soil confluence module, a surface confluence module, a groundwater confluence module, and a river confluence module. The soil confluence module is a three-dimensional calculation module that considers water exchange in both vertical and horizontal directions. When the WRF-Hydro mode is running, the land surface module on the coarse grid is first calculated, and then the confluence module on the fine grid is calculated, including the soil confluence module, the surface confluence module, the groundwater confluence module, and the river confluence module. In the WRF-Hydro mode, the global reanalysis data FNL is used as the initial field for the WRF-Hydro mode operation, and FNL is input into the WRF mode to obtain the initial field state of the WRF-Hydro mode operation; Step 2: First, collect the meteorological driving data, underlying surface data, and river network data of the watershed to be simulated as model input data, and perform bilinear interpolation on the obtained GLDAS data to obtain the initial driving data of the watershed to be simulated; Step 3: First, obtain precipitation data for the simulated basin from four precipitation products: CMFD, CMORPH, TRMM, and GLDAS. Then, compare the accuracy of the precipitation data from these four precipitation products in the simulated basin. Select the precipitation data with the best accuracy and combine it with the non-precipitation field from GLDAS to obtain the final driving data for the WRF-Hydro model. Step 4: Input the final driving data of the watershed to be simulated and the initial field data of the model operation into the WRF-Hydro model, then run the WRF-Hydro model. The model land surface module calculates the runoff of the input relevant water body data, and then downscales the surface runoff and soil moisture data to the confluence grid through the grid for calculation, and finally obtains the runoff of the watershed to be simulated.

2. The method for simulating and calculating runoff in the source region of the Yellow River according to claim 1, characterized in that: In step 1, the temporal and spatial resolutions of the WRF-Hydro model driving data are 3 h and 5.0 km × 5.0 km, respectively, the confluence grid resolution is 500.0 m × 500.0 m, the calculation time step is set to 20 s, and the runoff simulation results are output hourly.

3. The method for simulating and calculating runoff in the source region of the Yellow River according to claim 1, characterized in that: In step 2, the meteorological driving data consists of seven variables: downward longwave radiation, downward shortwave radiation, surface air pressure, specific humidity, air temperature, near-surface wind speed, and precipitation rate. The temporal and spatial resolutions of the GLDAS data are 3 hours and 0.25°×0.25°, respectively.

4. The method for simulating and calculating runoff in the source region of the Yellow River according to claim 1, characterized in that: In step 2, the underlying surface data are all from the WRF pre-processing system, and the river network data are from hydrological data and maps based on multi-scale elevation derivatives. Elevation data with a resolution of 90.0m×90.0m are selected, and accurate river network information is extracted.

5. The method for simulating and calculating runoff in the source region of the Yellow River according to claim 1, characterized in that: In step three, when performing accuracy comparison, the precipitation at the stations in the basin to be simulated is first interpolated into 0.25°×0.25° grid data using the Cressman objective analysis method. Then, bilinear interpolation is used to interpolate the four types of precipitation products, CMFD, CMORPH, TRMM, and GLDAS, to the same resolution. Then, spatial accuracy comparison is performed using the grid data, and the bilinear interpolation method is further used to interpolate the grid data to the stations. The arithmetic mean of the stations is used to represent the precipitation in the basin to be simulated, and the temporal accuracy of the precipitation products is compared.

6. The method for simulating and calculating runoff in the source region of the Yellow River according to claim 1, characterized in that: In step three, four evaluation indicators, namely correlation coefficient, root mean square error, Nash coefficient and relative deviation, are selected to judge the accuracy of precipitation. At the same time, three statistical indicators, namely detection rate, false alarm rate and detection success rate, are used to evaluate the ability of different precipitation products to capture precipitation in the Yellow River source area.

7. The method for simulating and calculating runoff in the source region of the Yellow River according to claim 1, characterized in that: In step 4, the final driving data of the watershed to be simulated are precipitation data from CMFD and non-precipitation data from GLDAS, and the initial field data required by the WRF-Hydro model are provided by FNL data.

Images