Future rain and snow change simulation method based on statistical downscaling and wet bulb temperature model

Abstract

The invention discloses a future rain and snow change simulation method based on statistical downscaling and a wet bulb temperature model. The method is specifically implemented according to the following steps: step 1, collecting meteorological data of a research area in a historical period and a future period; step 2, correcting meteorological data under various climate modes and different emission scenes in a future period; step 3, calculating day-by-day wet bulb temperature in a future period; 4, calculating three empirical parameters and two threshold temperatures in the daily time period; and step 5, dividing daily rainfall types in the future period. According to the invention, the problems of low rainfall type identification precision and limited rain and snow prediction capability under complex conditions in the existing rain and snow segmentation algorithm are solved.

Description

Technical Field

[0001] The present invention belongs to the cross - technical field of statistics, physics and meteorology, and particularly relates to a method for simulating future rain and snow changes based on statistical downscaling and wet - bulb temperature models. Background Technique

[0002] Precipitation refers to solid water, liquid water or a mixture of solid and liquid water that falls from the atmosphere to the ground. As an important indicator for monitoring climate change, precipitation is a key factor reflecting the changes in the atmospheric water cycle and its impact on regional climate and water resources. Precipitation is mainly divided into three types according to different phases: rainfall, snowfall and sleet. There are significant differences in the impact of different precipitation types on energy flow and surface runoff. When the precipitation type changes from snowfall to rainfall, it will lead to the earlier occurrence of spring runoff and at the same time reduce the water resource reserves in summer. Different precipitation types also have very different effects on the surface albedo, thus changing the energy balance of the local earth - atmosphere system. In addition, the change of precipitation type will directly affect the change of glacier mass balance.

[0003] Therefore, under the background of global warming, it is necessary to distinguish and simulate different precipitation types, predict the changing trend of rain and snow over time in the future, and thus provide a scientific reference basis for aspects such as the response to climate change, the monitoring and early warning of different types of natural disasters, and the formulation of coping strategies.

[0004] Currently, the existing rain - snow segmentation algorithms for distinguishing precipitation types mainly include representative methods such as the KS model, the double - critical temperature model, and the wet - bulb temperature model. The KS model (Koistinen J, Saltikoff E. Experience of customer products of accumulated snow, sleet and rain[J]. COST75 Advanced Weather Radar Systems, 1998, 397:406.) mainly calculates the occurrence probability of rainfall events in precipitation types through air temperature and relative humidity. However, since this model does not consider the existence of mixed solid - liquid precipitation, that is, sleet type, it is not comprehensive enough when judging precipitation types and has limitations.

[0005] The double - critical temperature model (Chen Rensheng, Lü Shihua, Kang Ersi, etc. Distributed hydro - thermal coupling model for inland river alpine mountainous basins (Ⅰ): Model principle[J]. Advances in Earth Science, 2006, 21(8):806.) mainly uses the maximum critical temperature and the minimum critical temperature to compare with the daily average temperature to divide precipitation types. However, since this model only relies on a single meteorological data of air temperature and does not consider other meteorological factors affecting precipitation types, its applicability and accuracy are relatively low under complex meteorological conditions.

[0006] The wet-bulb temperature model (Ding B, Yang K, Qin J, et al. The dependence of precipitation types on surface elevation and meteorological conditions and its parameterization [J]. Journal of hydrology, 2014, 513: 154 - 163.) includes six meteorological indicators such as daily average wet-bulb temperature, air temperature, and relative humidity when discriminating precipitation types. Compared with most models that judge precipitation types based on a single meteorological index, the wet-bulb temperature model has higher reliability and accuracy in alpine regions, especially in the Tibetan Plateau where snowfall is concentrated and meteorological stations are limited. However, due to the high precision requirements of the model for input data, especially in alpine mountainous areas, more advanced processing methods need to be adopted for the acquisition and quality control of meteorological data to improve the spatial resolution of the data and thus enhance the simulation accuracy of the model. Summary of the Invention

[0007] The object of the present invention is to provide a method for simulating future rain and snow changes based on statistical downscaling and the wet-bulb temperature model, which solves the problems of low accuracy in identifying precipitation types in existing rain and snow segmentation algorithms and limitations in rain and snow prediction capabilities under complex conditions.

[0008] The technical solution adopted by the present invention is a method for simulating future rain and snow changes based on statistical downscaling and the wet-bulb temperature model, which is specifically implemented according to the following steps:

[0009] Step 1: Collect meteorological data in the historical period and future period of the study area;

[0010] Step 2: Calibrate meteorological data under different emission scenarios of multiple climate models in the future period;

[0011] Step 3: Calculate the daily wet-bulb temperature in the future period;

[0012] Step 4: Calculate three empirical parameters and two threshold temperatures in the daily time period;

[0013] Step 5: Divide the daily precipitation types in the future period.

[0014] The present invention is also characterized in that,

[0015] Step 1 is specifically implemented according to the following steps:

[0016] The daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data statistically collected from meteorological stations and remote sensing products covering the study area are used as historical period observation data, and the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data from multiple global climate models in the Sixth Coupled Model Intercomparison Project CMIP6 under three emission scenarios of SSP126, SSP245, and SSP585 are used as future period simulation data.

[0017] Step 2 is specifically implemented according to the following steps:

[0018] Based on the relevant data collected in Step 1, and using the historical observations and historical simulation values of the CMIP6 global climate models, the future meteorological simulation data under different emission scenarios of CMIP6 are corrected through a statistical downscaling method.

[0019] In Step 2, the equal-distance cumulative distribution function matching method EDCDF is used as the statistical downscaling method to correct the meteorological data under different emission scenarios of multiple climate models in the future period. The specific method is as follows:

[0020]

[0021] x correct = x + Δ

[0022] Δ represents the difference correction amount between the historical period observation data and the simulation data; x represents climate elements, including average temperature, rainfall, relative humidity, and atmospheric pressure; x correct represents the corrected climate element; F represents the cumulative probability distribution function; oc represents the historical period observation value; mc represents the historical period simulation value; ms represents the future period simulation value.

[0023] Step 3 is specifically implemented according to the following steps:

[0024] Using the corrected future daily average temperature, relative humidity, and atmospheric pressure data in Step 2, the daily wet-bulb temperature data under different future emission scenarios are calculated through the formula as follows:

[0025]

[0026] T w represents the daily average wet-bulb temperature, with the unit of °C; T a represents the average temperature, with the unit of °C; RH represents the relative humidity, with the unit of %; p s represents the atmospheric pressure, with the unit of hPa; e sat (T a ) represents the saturation vapor pressure at temperature T a .

[0027] Step 4 is specifically implemented according to the following steps:

[0028] Based on the daily meteorological data corrected in Step 2 and the daily wet-bulb temperatures calculated in Step 3, calculate the empirical parameters of the three wet-bulb temperature models. Then, use the empirical parameters to calculate the minimum threshold temperature during rainfall and the maximum threshold temperature during snowfall. The method for calculating the three empirical parameters and the two threshold temperatures in the daily time period is as follows:

[0029] ΔT = 0.215 - 0.099×RH + 1.018×RH 2

[0030] ΔS = 2.374 - 1.634×RH

[0031] T0 = -5.87 - 0.1042×Z + 0.0885×Z² + 16.06×RH - 9.614×RH 2

[0032] Among them, ΔT, ΔS, and T0 are all empirical parameters used to calculate the subsequent two threshold temperatures;

[0033] Z represents the elevation, with the unit of km; T min represents the minimum threshold temperature during rainfall, with the unit of °C; T max represents the maximum threshold temperature during snowfall, with the unit of °C.

[0034] Step 5 is specifically implemented according to the following steps:

[0035] Using the wet-bulb temperatures calculated in Step 2 and the two threshold temperatures calculated in Step 4, by comparing them with each other, divide the precipitation into three forms: rainfall, snowfall, and sleet, and simulate the rain and snow changes under different emission scenarios in the future period, specifically as follows:

[0036]

[0037] P type represents the precipitation type, including rainfall, snowfall, and sleet; Snow represents the snowfall event; Sleet represents the sleet event; Rain represents the rainfall event, and T w represents the daily average wet-bulb temperature, with the unit of °C; T min represents the minimum threshold temperature during rainfall, with the unit of °C; T max represents the maximum threshold temperature during snowfall, with the unit of °C.

[0038] The beneficial effects of the present invention are as follows: 1. By using the equidistant cumulative distribution function matching method and based on the observed data in historical periods, the present invention performs statistical downscaling correction on the input meteorological data in future periods, realizing the information conversion of meteorological data from large scales to small scales, greatly improving the spatial resolution and accuracy of the input data, and thus enhancing the meteorological prediction ability and feature capture accuracy for local regions. 2. Based on the wet-bulb temperature model and combined with the statistical downscaling method, the present invention solves the problem in existing rain-snow segmentation algorithms that it is impossible to accurately distinguish precipitation types, especially the difficulty in judging solid-liquid mixed precipitation (sleet events). At the same time, compared with existing rain-snow segmentation algorithms, the present invention effectively improves the application performance of the model in alpine mountainous areas where meteorological data is scarce and difficult to collect, and enhances the applicability of the model under complex climate conditions. 3. By combining the future simulation data of multiple global climate models in CMIP6 and considering the changes in meteorological data under three emission scenarios, namely SSP126, SSP245, and SSP585, the present invention effectively reduces the uncertainty under a single future prediction scenario, provides more comprehensive and diverse prediction results of rain-snow changes, and provides a scientific reference basis for policy-making and risk assessment in response to climate change.

[0039] In summary, by combining the statistical downscaling method and the wet-bulb temperature model, the present invention solves the technical problems of low accuracy in identifying precipitation types in existing rain-snow segmentation algorithms and limited rain-snow prediction ability under complex conditions. The present invention has the characteristics of high accuracy, comprehensive results for future rain-snow change simulation, especially in regions with scarce meteorological data such as alpine mountainous areas, with stronger prediction ability and wide applicability, and is easy to promote and apply. Brief Description of the Drawings

[0040] Figure 1 It is a schematic diagram of the relationship between precipitation types and the average daily wet-bulb temperature (T w ) in the wet-bulb temperature model;

[0041] Figure 2 It is a flow chart of the future rain-snow change simulation method based on statistical downscaling and the wet-bulb temperature model;

[0042] Figure 3 It is a model verification diagram of the wet-bulb temperature model, historical observed values, and two other representative rain-snow segmentation algorithms;

[0043] Figure 4 It is a simulation result diagram taking the rain-snow change of Huma Station from 2015 to 2100 as an example. Detailed Embodiment

[0044] The present invention will be described in detail below in conjunction with the drawings and specific embodiments.

[0045] The future rain and snow change simulation method based on statistical downscaling and wet-bulb temperature model is specifically implemented according to the following steps:

[0046] Step 1: Collect meteorological data in the historical period and future period of the study area;

[0047] Step 1 is specifically implemented according to the following steps:

[0048] Select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data statistically calculated by meteorological stations and remote sensing products covering the study area as historical period observation data, and select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data of multiple global climate models in the sixth international coupled comparison project CMIP6 under three emission scenarios SSP126, SSP245, and SSP585 as future period simulation data.

[0049] Step 2: Correct the meteorological data of multiple climate models under different emission scenarios in the future period;

[0050] Step 2 is specifically implemented according to the following steps:

[0051] Based on the relevant data collected in Step 1, and based on the historical observation values and historical simulation values of the CMIP6 global climate model, the future meteorological simulation data under different emission scenarios of CMIP6 is corrected through the statistical downscaling method.

[0052] In Step 2, the statistical downscaling method for correcting the meteorological data of multiple climate models under different emission scenarios in the future period adopts the equal-distance cumulative distribution function matching method EDCDF. The specific method is as follows:

[0053]

[0054] x correct = x + Δ

[0055] Δ represents the difference correction amount between the historical period observation data and the simulation data; x represents the climate elements, including average temperature, rainfall, relative humidity, and atmospheric pressure; x correct represents the corrected climate element; F represents the cumulative probability distribution function; oc represents the historical period observation value; mc represents the historical period simulation value; ms represents the future period simulation value.

[0056] Step 3: Calculate the daily wet-bulb temperature in the future period;

[0057] Step 3 is specifically implemented according to the following steps:

[0058] Using the corrected future daily average temperature, relative humidity, and atmospheric pressure data in Step 2, the daily wet-bulb temperature data under different future emission scenarios is calculated through the formula as:

[0059]

[0060] T w represents the daily average wet-bulb temperature, with the unit of °C; T a represents the average air temperature, with the unit of °C; RH represents the relative humidity, with the unit of %; p s represents the atmospheric pressure, with the unit of hPa; e sat (T a ) represents the saturation vapor pressure at the temperature of T a when the temperature is T

[0061] Step 4: Calculate three empirical parameters and two threshold temperatures for the daily time period;

[0062] Step 4 is specifically implemented according to the following steps:

[0063] According to the daily meteorological data corrected in Step 2 and the daily wet-bulb temperature calculated in Step 3, calculate three empirical parameters of the wet-bulb temperature model. Then, use the empirical parameters to calculate the lowest threshold temperature when rainfall occurs and the highest threshold temperature when snowfall occurs. The method for calculating the three empirical parameters and two threshold temperatures for the daily time period is as follows:

[0064] ΔT = 0.215 - 0.099×RH + 1.018×RH 2

[0065] ΔS = 2.374 - 1.634×RH

[0066] T0 = -5.87 - 0.1042×Z + 0.0885×Z 2 + 16.06×RH - 9.614×RH 2

[0067]

[0068] where ΔT, ΔS, and T0 are all empirical parameters used to calculate the following two threshold temperatures;

[0069] Z represents the elevation, with the unit of km; T min represents the lowest threshold temperature when rainfall occurs, with the unit of °C; T max represents the highest threshold temperature when snowfall occurs, with the unit of °C.

[0070] Step 5: Classify the daily precipitation types in the future period.

[0071] Step 5 is specifically implemented according to the following steps:

[0072] Using the wet-bulb temperature calculated in Step 2 and the two threshold temperatures calculated in Step 4, by comparing them with each other, precipitation is divided into three forms: rainfall, snowfall, and sleet, and the rain-snow changes under different emission scenarios in the future period are simulated as follows:

[0073]

[0074] P type represents the precipitation type, including rainfall, snowfall, and sleet; Snow represents a snowfall event; Sleet represents a sleet event; Rain represents a rainfall event, T w represents the daily average wet-bulb temperature, with the unit of °C; T min represents the lowest threshold temperature when rainfall occurs, with the unit of °C; T max represents the highest threshold temperature when snowfall occurs, with the unit of °C.

[0075] The present invention mainly uses the statistical downscaling method to complete the correction of daily meteorological data, combines the wet-bulb temperature model to achieve the classification of precipitation types, and thus simulates the rain-snow change trend in the future period. This method solves the technical problems such as fuzzy identification of precipitation types and limitations in predicting rain-snow changes under complex conditions in the existing rain-snow segmentation algorithms by accurately classifying precipitation types, improving the spatial resolution of meteorological data, and considering various emission scenario simulations, and provides a scientific reference basis for policy-making, risk assessment in response to climate change, and water resource management, etc.

[0076] Example 1

[0077] The method for simulating future rain-snow changes based on statistical downscaling and wet-bulb temperature model of the present invention is specifically implemented according to the following steps:

[0078] Step 1: Collect the meteorological data of the study area in the historical period and the future period;

[0079] Step 2: Correct the meteorological data under different emission scenarios of multiple climate models in the future period;

[0080] Step 3: Calculate the daily wet-bulb temperature in the future period;

[0081] Step 4: Calculate three empirical parameters and two threshold temperatures in the daily time period;

[0082] Step 5: Classify the daily precipitation types in the future period.

[0083] Example 2

[0084] The method for simulating future rain-snow changes based on statistical downscaling and wet-bulb temperature model of the present invention is specifically implemented according to the following steps:

[0085] Step 1: Collect meteorological data for the historical and future periods in the study area;

[0086] Step 1 is specifically implemented according to the following steps:

[0087] Select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data statistically obtained from meteorological stations and remote sensing products covering the study area as the historical period observation data, and select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data of multiple global climate models in the Sixth Coupled Model Intercomparison Project CMIP6 under three emission scenarios, SSP126, SSP245, and SSP585, as the future period simulation data.

[0088] Step 2: Correct the meteorological data of multiple climate models under different emission scenarios in the future period;

[0089] Step 3: Calculate the daily wet-bulb temperature in the future period;

[0090] Step 4: Calculate three empirical parameters and two threshold temperatures for the daily time period;

[0091] Step 5: Classify the daily precipitation types in the future period.

[0092] Example 3

[0093] The future rain and snow change simulation method of the present invention based on statistical downscaling and wet-bulb temperature model is specifically implemented according to the following steps:

[0094] Step 1: Collect meteorological data for the historical and future periods in the study area;

[0095] Step 1 is specifically implemented according to the following steps:

[0096] Select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data statistically obtained from meteorological stations and remote sensing products covering the study area as the historical period observation data, and select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data of multiple global climate models in the Sixth Coupled Model Intercomparison Project CMIP6 under three emission scenarios, SSP126, SSP245, and SSP585, as the future period simulation data.

[0097] Step 2: Correct the meteorological data of multiple climate models under different emission scenarios in the future period;

[0098] Step 2 is specifically implemented according to the following steps:

[0099] According to the relevant data collected in Step 1, based on the historical observations and the historical simulation values of the CMIP6 global climate models, the future meteorological simulation data under different emission scenarios of CMIP6 are corrected by the statistical downscaling method.

[0100] In Step 2, the statistical downscaling method for correcting meteorological data under different emission scenarios of multiple climate models in the future period adopts the equal-distance cumulative distribution function matching method EDCDF. The specific method is as follows:

[0101]

[0102] x correct = x + Δ

[0103] Δ represents the difference correction amount between the observed data and the simulated data in the historical period; x represents climate elements, including average temperature, rainfall, relative humidity, and atmospheric pressure; x correct represents the corrected climate element; F represents the cumulative probability distribution function; oc represents the observed value in the historical period; mc represents the simulated value in the historical period; ms represents the simulated value in the future period.

[0104] Step 3: Calculate the daily wet-bulb temperature in the future period;

[0105] Step 4: Calculate three empirical parameters and two threshold temperatures in the daily time period;

[0106] Step 5: Classify the daily precipitation types in the future period.

[0107] Example 4

[0108] The future rain and snow change simulation method based on statistical downscaling and wet-bulb temperature model of the present invention is specifically implemented according to the following steps:

[0109] Step 1: Collect the meteorological data of the historical period and the future period in the study area;

[0110] Step 1 is specifically implemented according to the following steps:

[0111] Select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data statistically obtained from the meteorological stations and remote sensing products covering the study area as the observed data in the historical period, and select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data of multiple global climate models in the sixth international coupled comparison project CMIP6 under three emission scenarios SSP126, SSP245, and SSP585 as the simulated data in the future period.

[0112] Step 2: Correct the meteorological data under different emission scenarios of multiple climate models in the future period;

[0113] Step 2 is specifically implemented according to the following steps:

[0114] Based on the relevant data collected in Step 1, and based on historical observations and historical simulation values of CMIP6 global climate models, the future meteorological simulation data under different emission scenarios of CMIP6 are corrected through a statistical downscaling method.

[0115] In Step 2, the statistical downscaling method for correcting meteorological data under different emission scenarios of multiple climate models in the future adopts the Equal-Distance Cumulative Distribution Function Matching method (EDCDF). The specific method is as follows:

[0116]

[0117] x correct = x + Δ

[0118] Δ represents the difference correction amount between the observed data and the simulated data in the historical period; x represents climate elements, including average temperature, rainfall, relative humidity, and atmospheric pressure; x correct represents the corrected climate element; F represents the cumulative probability distribution function; oc represents the observed value in the historical period; mc represents the simulated value in the historical period; ms represents the simulated value in the future period.

[0119] Step 3: Calculate the daily wet-bulb temperature in the future period;

[0120] Step 3 is specifically implemented according to the following steps:

[0121] Using the corrected future daily average temperature, relative humidity, and atmospheric pressure data in Step 2, the daily wet-bulb temperature data under different future emission scenarios are calculated through the formula as follows:

[0122]

[0123] T w represents the daily average wet-bulb temperature, with the unit of °C; T a represents the average temperature, with the unit of °C; RH represents the relative humidity, with the unit of %; p s represents the atmospheric pressure, with the unit of hPa; e sat (T a ) represents the saturated vapor pressure at the temperature of T a

[0124] Step 4: Calculate three empirical parameters and two threshold temperatures in the daily time period;

[0125] Step 5: Divide the daily precipitation type in the future period.

[0126] Example 5

[0127] The future rain and snow change simulation method of the present invention based on statistical downscaling and wet-bulb temperature model is specifically implemented according to the following steps: ​

[0128] Step 1: Collect meteorological data for the historical and future periods in the study area;

[0129] Step 1 is specifically implemented according to the following steps:

[0130] Select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data statistically obtained from meteorological stations and remote sensing products covering the study area as the historical period observation data, and select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data of multiple global climate models in the sixth International Coupled Model Intercomparison Project CMIP6 under three emission scenarios, SSP126, SSP245, and SSP585, as the future period simulation data.

[0131] Step 2: Correct the meteorological data of multiple climate models under different emission scenarios in the future period;

[0132] Step 2 is specifically implemented according to the following steps:

[0133] Based on the relevant data collected in Step 1, and based on the historical observations and historical simulation values of the CMIP6 global climate models, the future meteorological simulation data under different emission scenarios of CMIP6 are corrected through a statistical downscaling method.

[0134] In Step 2, the statistical downscaling method for correcting the meteorological data of multiple climate models under different emission scenarios in the future period uses the equal-distance cumulative distribution function matching method EDCDF. The specific method is as follows:

[0135]

[0136] x correct = x + Δ

[0137] Δ represents the difference correction amount between the historical period observation data and the simulation data; x represents climate elements, including average temperature, rainfall, relative humidity, and atmospheric pressure; x correct represents the corrected climate element; F represents the cumulative probability distribution function; oc represents the historical period observation value; mc represents the historical period simulation value; ms represents the future period simulation value.

[0138] Step 3: Calculate the daily wet-bulb temperature in the future period;

[0139] Step 3 is specifically implemented according to the following steps:

[0140] Using the corrected future daily average temperature, relative humidity, and atmospheric pressure data in Step 2, the daily wet-bulb temperature data under different future emission scenarios are calculated through the formula as follows:

[0141]

[0142] T w represents the daily average wet-bulb temperature, with the unit of °C; T a represents the average temperature, with the unit of °C; RH represents the relative humidity, with the unit of %; p s represents the atmospheric pressure, with the unit of hPa; e sat (T a ) represents the saturated vapor pressure at the temperature of T a when it is.

[0143] Step 4: Calculate three empirical parameters and two threshold temperatures for the daily time period;

[0144] Step 4 is specifically implemented according to the following steps:

[0145] Based on the daily meteorological data corrected in Step 2 and the daily wet-bulb temperature calculated in Step 3, calculate three empirical parameters of the wet-bulb temperature model. Then, use the empirical parameters to calculate the lowest threshold temperature when rainfall occurs and the highest threshold temperature when snowfall occurs. The method for calculating the three empirical parameters and the two threshold temperatures for the daily time period is as follows:

[0146] ΔT = 0.215 - 0.099×RH + 1.018×RH 2

[0147] ΔS = 2.374 - 1.634×RH

[0148] T0 = -5.87 - 0.1042×Z + 0.0885×Z 2 + 16.06×RH - 9.614×RH 2

[0149]

[0150]

[0151] Among them, ΔT, ΔS, and T0 are all empirical parameters used to calculate the subsequent two threshold temperatures;

[0152] Z represents the altitude, with the unit of km; T min represents the lowest threshold temperature when rainfall occurs, with the unit of °C; T max represents the highest threshold temperature when snowfall occurs, with the unit of °C.

[0153] Step 5: Divide the daily precipitation types for the future period.

[0154] Example 6

[0155] Combined with a future rain and snow change simulation method based on statistical downscaling and wet-bulb temperature model in the present invention, the following example is given: Taking Huma Station in Heilongjiang Province as an example, the period from 1980 to 2014 is taken as the historical period. Through historical meteorological observation values and simulation values, the meteorological simulation values in the future period are processed by the statistical downscaling method. The meteorological data mainly includes variables such as air temperature, air pressure, precipitation, and relative humidity. Combining with the wet-bulb temperature model, the rain and snow changes at the station in the future period from 2015 to 2100 are simulated. The method includes the following 6 steps.

[0156] Step 1: Collect relevant data. It includes the following 3 sub-steps:

[0157] (1.1) Taking Huma County in Heilongjiang Province as the research area. The observed meteorological data in the historical period (1980 - 2014) mainly comes from the Huma meteorological station of the National Meteorological Information Center of the China Meteorological Administration (https: / / data.cma.cn / ). The data mainly includes variables such as air temperature, air pressure, precipitation, and relative humidity, and the time resolution is daily.

[0158] (1.2) The simulated meteorological data in the future period (2015 - 2100) comes from 3 global climate models in CMIP6, namely IPSL-CM6A-LR, MIROC6, and MRI-ESM2-0. These models have good performance in the relevant research on existing rain and snow changes. The meteorological data is the same as the variables taken in the historical period, mainly including variables such as air temperature, air pressure, precipitation, and relative humidity, and the time resolution is daily. To ensure the comprehensiveness and reliability of the prediction results, three different emission scenarios SSP126, SSP245, and SSP585 under CMIP6 are selected.

[0159] (1.3) Collect the ground elevation data of the study area, and adopt the digital elevation model with a spatial resolution of 30m provided by SRTM.

[0160] Step 2: Correct the meteorological simulation data in the future period based on the statistical downscaling method. It includes the following 2 sub-steps:

[0161] (2.1) Clip the observed values and simulation values of the historical period meteorological data (average air temperature, rainfall, relative humidity, atmospheric pressure), as well as the simulation values under the three emission scenarios in the future period, with Huma County in Heilongjiang Province as the research area, and unify the spatial resolution of the meteorological data to 0.1°×0.1° (about 11km).

[0162] (2.2) To improve the accuracy of simulated data in future periods, the equal-distance cumulative distribution function method (EDCDF) is used for statistical downscaling. The cumulative probability distribution function between the observed values and simulated values of meteorological data in historical periods is established, and it is assumed that under this cumulative probability, the difference between the observed values and simulated values remains the same in future periods. Then, the meteorological data in future periods are corrected according to the following formula:

[0163]

[0164] x correct = x + Δ

[0165] x: Climate elements, including average temperature, rainfall, relative humidity, atmospheric pressure, etc.;

[0166] x correct : Corrected climate elements;

[0167] F: Cumulative probability distribution function;

[0168] oc: Observed values in historical periods;

[0169] mc: Simulated values in historical periods;

[0170] ms: Simulated values in future periods.

[0171] Step 3: Calculate the daily wet-bulb temperature of different global climate models under three emission scenarios in future periods:

[0172] Based on the meteorological data such as average temperature, relative humidity, and atmospheric pressure in future periods after the statistical downscaling in Step 2, calculate the daily average wet-bulb temperature of the global climate models IPSL-CM6A-LR, MIROC6, and MRI-ESM2-0 under the SSP126, SSP245, and SSP585 emission scenarios respectively according to the following formula:

[0173]

[0174] T w : Daily average wet-bulb temperature (°C);

[0175] T a : Average temperature (°C);

[0176] RH: Relative humidity (%);

[0177] p s : Atmospheric pressure (hPa);

[0178] e sat (T a ): Saturation vapor pressure at temperature T a .

[0179] Step 4. Calculate three empirical parameters and two threshold temperatures for different global climate models in three emission scenarios in the future period with a daily time resolution, including the following two sub-steps:

[0180] (4.1) Based on the relative humidity data and ground elevation data in the future period after the statistical downscaling process in Step 2, calculate the three empirical parameters at the daily time scale according to the following formula:

[0181] ΔT = 0.215 - 0.099×RH + 1.018×RH 2

[0182] ΔS = 2.374 - 1.634×RH

[0183] T0 = -5.87 - 0.1042×Z + 0.0885×Z 2 + 16.06×RH - 9.614×RH 2

[0184] ΔT, ΔS, T0: Empirical parameters used to calculate the following two threshold temperatures;

[0185] RH: The same meaning as above;

[0186] Z: Elevation (km).

[0187] (4.2) Using the three empirical parameters calculated in Step (4.1), further determine the minimum threshold temperature at the time of rainfall and the maximum threshold temperature at the time of snowfall for the IPSL-CM6A-LR, MIROC6, and MRI-ESM2-0 global climate models under the SSP126, SSP245, and SSP585 emission scenarios respectively at the daily time scale according to the following formula:

[0188]

[0189] T min : The minimum threshold temperature at the time of rainfall (°C);

[0190] T max : The maximum threshold temperature at the time of snowfall (°C);

[0191] ΔT, ΔS, T0: The same meaning as above.

[0192] Step 5. Divide the daily precipitation types for different global climate models in three emission scenarios in the future period:

[0193] Based on the wet-bulb temperature, the minimum threshold temperature at the time of rainfall, and the maximum threshold temperature at the time of snowfall calculated in Steps 3 and 4 respectively, divide the precipitation types in the future period according to the following formula:

[0194]

[0195] P type : Precipitation type, including rainfall, snowfall, and sleet;

[0196] Snow: Snowfall event;

[0197] Sleet: Sleet event;

[0198] Rain: Rainfall event;

[0199] T w : The same meaning as above;

[0200] T min : The same meaning as above;

[0201] T max : The same meaning as above.

[0202] Step 6: Based on the daily rainfall data for the future period after statistical downscaling in Step 2, combined with the judgment of precipitation type in Step 5, the daily spatial distributions of rainfall events, snowfall events, and sleet events under the SSP1-2.6, SSP2-4.5, and SSP5-8.5 emission scenarios of the IPSL-CM6A-LR, MIROC6, and MRI-ESM2-0 global climate models are obtained, and then the rain and snow changes at the stations in the future period are analyzed.

[0203] Figure 3 It is a model validation diagram of the wet-bulb temperature model with historical observations and two other representative rain-snow segmentation algorithms. To verify the simulation accuracy of the wet-bulb temperature model, the snowfall observations recorded at the Huma meteorological station in 1980 were used. The wet-bulb temperature model and two other representative models in the current rain-snow division algorithms, the KS model and the double critical temperature model, were used for rain-snow division, and the daily snowfall amounts calculated by different models were compared. The KS model calculates different types of precipitation amounts as follows:

[0204]

[0205] P(Rain): Probability of rainfall, with a value between 0 and 1. The closer its value is to 1, the more likely a rainfall event is to occur; conversely, the closer it is to 0, the greater the likelihood of snowfall.

[0206] T: The same meaning as above;

[0207] RH: The same meaning as above.

[0208] The double critical temperature model calculates different types of precipitation amounts as follows:

[0209]

[0210] P L : The daily rainfall (mm) judged by the double critical temperature model;

[0211] P: Daily rainfall (mm);

[0212] T: The same meaning as above;

[0213] T L : The highest critical temperature (°C), set as 2 °C;

[0214] T S : The lowest critical temperature (°C), set as 0 °C.

[0215] The figure shows a comparison chart of the result accuracy between the simulated values of three models and the observed values at meteorological stations. From Figure 3 it can be seen that the fitting accuracy between the simulated result of the wet-bulb temperature model adopted by the present invention and the observed result is higher.

[0216] Figure 4 It is a simulated result chart taking the rain and snow changes in Huma Station from 2015 to 2100 as an example. It can be clearly seen in the figure the future change trends of different types of precipitation (rainfall, snowfall and sleet) under different emission scenarios.

[0217] Based on the wet-bulb temperature model, this method adopts the statistical downscaling method to correct the input meteorological data, combines multiple global climate models in the Sixth Coupled Model Intercomparison Project (CMIP6), and considers the changes of meteorological factors under three different emission scenarios (sustainable development scenario SSP126, historical development scenario SSP245 and fossil fuel-dependent development scenario SSP585), so as to be able to more accurately distinguish precipitation types and simulate and predict the rain and snow change trends in the future period.

Claims

1. A method for simulating future rain and snow changes based on statistical downscaling and wet-bulb temperature model, characterized in that, The implementation is specifically carried out according to the following steps: Step 1: Collect the meteorological data of the study area in historical and future periods; Step 2: Calibrate the meteorological data under different emission scenarios of multiple climate models in the future period; Step 3: Calculate the daily wet-bulb temperature in the future period; Step 4: Calculate three empirical parameters and two threshold temperatures in the daily time period; Step 5: Classify the daily precipitation types in the future period.

2. The future rain and snow change simulation method based on statistical downscaling and wet-bulb temperature model according to claim 1, wherein The specific implementation of Step 1 is carried out according to the following steps: Select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data statistically obtained from meteorological stations and remote sensing products covering the study area as the historical period observation data, and select the daily average temperature, precipitation, relative humidity, and atmospheric pressure meteorological data of multiple global climate models in the Sixth Coupled Model Intercomparison Project CMIP6 under three emission scenarios SSP126, SSP245, and SSP585 as the future period simulation data.

3. The method for simulating future rain and snow changes based on statistical downscaling and wet-bulb temperature model according to claim 2, wherein The specific implementation of Step 2 is carried out according to the following steps: Based on the relevant data collected in Step 1, and based on the historical observation values and the historical simulation values of the CMIP6 global climate model, calibrate the future meteorological simulation data under different emission scenarios of CMIP6 through a statistical downscaling method.

4. The method for simulating future rain and snow changes based on statistical downscaling and wet-bulb temperature model according to claim 3, wherein In Step 2, the statistical downscaling method for calibrating the meteorological data under different emission scenarios of multiple climate models in the future period adopts the Equal-Distance Cumulative Distribution Function Matching method EDCDF. The specific method is as follows: x correct = x + Δ △ represents the difference correction amount between the observed data and the simulated data in the historical period; x represents climate elements, including average temperature, rainfall, relative humidity, and atmospheric pressure; x correct represents the corrected climate element; F represents the cumulative probability distribution function; oc represents the observed value in the historical period; mc represents the simulated value in the historical period; ms represents the simulated value in the future period.

5. The future rain and snow change simulation method based on statistical downscaling and wet-bulb temperature model according to claim 4, characterized in that The specific implementation of Step 3 is carried out according to the following steps: Using the calibrated future daily average temperature, relative humidity, and atmospheric pressure data in Step 2, calculate the daily wet-bulb temperature data under different future emission scenarios through the formula: T w represents the daily average wet-bulb temperature, with the unit of °C; T a represents the average temperature, with the unit of °C; RH represents the relative humidity, with the unit of %; p s represents the atmospheric pressure, with the unit of hPa; e sat (T a ) represents the saturated vapor pressure at the temperature of T a °C.

6. The method for simulating future rain and snow changes based on statistical downscaling and wet bulb temperature model according to claim 5, wherein The specific implementation of Step 4 is carried out according to the following steps: According to the daily meteorological data calibrated in Step 2 and the daily wet-bulb temperature calculated in Step 3, calculate three empirical parameters of the wet-bulb temperature model. Use the empirical parameters to further calculate the lowest threshold temperature when rainfall occurs and the highest threshold temperature when snowfall occurs. The method for calculating the three empirical parameters in the daily time period is as follows: ΔT = 0.215 - 0.099×RH + 1.018×RH 2 ΔS = 2.374 - 1.634×RH T0 = -5.87 - 0.1042×Z + 0.0885×Z 2 + 16.06×RH - 9.614×RH 2 。 7. The method for simulating future rain and snow changes based on statistical downscaling and wet bulb temperature model according to claim 6, wherein The calculation methods of the two threshold temperatures in Step 4 are as follows: Among them, ΔT, ΔS, and T0 are all empirical parameters, which are used to calculate the following two threshold temperatures; Z represents the elevation, with the unit of km; T min represents the lowest threshold temperature when rainfall occurs, with the unit of °C; T max represents the highest threshold temperature when snowfall occurs, with the unit of °C.

8. The method for simulating future rain and snow changes based on statistical downscaling and wet bulb temperature model according to claim 7, wherein The specific implementation of Step 5 is carried out according to the following steps: Using the wet-bulb temperature calculated in Step 2 and the two threshold temperatures calculated in Step 4, compare them with each other to classify the precipitation into three forms: rainfall, snowfall, and sleet, and simulate the rain and snow changes under different emission scenarios in the future period. The specific details are as follows: P type represents precipitation types, including rainfall, snowfall, and sleet; Snow represents a snowfall event; Sleet represents a sleet event; Rain represents a rainfall event, T w represents the daily average wet-bulb temperature, in °C; T min represents the lowest threshold temperature when rainfall occurs, in °C; T max represents the highest threshold temperature when snowfall occurs, in °C.

Images