A method for optimizing the layout of rainfall station networks in mountainous areas based on radar rainfall data

Abstract

The present invention discloses a method for optimizing the layout of rain gauge networks in mountainous areas based on radar rainfall data. The method comprises the following steps: Step 1: Collecting rainfall data and gridding the target area; Step 2: Identifying abnormal rain gauges; Step 3: Determining candidate locations for additional rain gauges; Step 4: Evaluating the accuracy and effectiveness of different rain gauge network optimization schemes; and Step 5: Determining the optimal rain gauge network layout scheme. The method, combined with radar rainfall data, effectively plans and optimizes rain gauge networks in complex mountainous terrain, thereby improving the accuracy and reliability of precipitation observations in mountainous watersheds and providing more accurate data support for flood disaster management and prevention in mountainous areas.

Description

Technical Field

[0001] The invention belongs to the technical field of hydrological and meteorological monitoring, and in particular relates to a method for optimizing the layout of a rainfall station network in a mountainous area based on radar rainfall measurement data. Background Art

[0002] As the impact of global climate change becomes increasingly significant, extreme weather events are becoming more frequent, particularly floods in mountainous watersheds, posing a serious threat to people's lives and property. Accurate and reliable rainfall monitoring is crucial for flood forecasting and early warning. Existing rain gauge deployment often relies on empirical judgment and limited meteorological observation data, failing to fully consider the impact of complex mountain terrain on rainfall distribution. The strong spatial heterogeneity and spatiotemporal variability of rainfall in mountainous areas can lead to blind spots in heavy rainfall observations, making it impossible to accurately capture localized heavy rainfall and rainstorm events, thus compromising the accuracy of flood forecasts and disaster warnings.

[0003] Currently, existing methods for optimizing rainfall station networks are mostly based on specific optimization algorithms, such as genetic algorithms and particle swarm optimization (PSO), which improve network coverage and monitoring accuracy by selecting optimal locations. However, most existing optimization methods rely on ground-based rainfall station data and ignore the potential of remote sensing technologies (such as radar rainfall measurement). Radar data has high temporal and spatial resolution and can provide more comprehensive precipitation monitoring information, but traditional methods fail to fully utilize this information. Therefore, although existing optimization methods can improve station network performance under certain conditions, they still face challenges in accuracy and efficiency in complex mountainous environments, and cannot achieve ideal monitoring results. Therefore, how to optimize traditional rainfall station networks in complex mountainous terrain by combining high-precision remote sensing data is a technical problem that needs to be solved urgently. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for optimizing the layout of a rainfall station network in a mountainous area based on radar rainfall measurement data, so as to solve the above-mentioned technical problems.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] The present invention discloses a method for optimizing the layout of a rainfall station network in a mountainous area based on radar rainfall measurement data, the method comprising the following steps:

[0007] Step 1: Collect rainfall data and divide the target area into grids: Collect the actual rainfall data P of all M rainfall stations in the target area. 雨量站 and P 雨量站 Radar rainfall data corresponding to the time series P 雷达 , divide the target area into the same grids according to the grid distribution of radar rainfall data, record a total of J, and assign a unique identifier to each grid;

[0008] Step 2: Identify abnormal rain gauges: Use the graph convolutional network model (GCN) to identify abnormal rain gauges. First, construct the node features of the rain gauges. If the actual rainfall data of each rain gauge is N hours, then use the node features of M rain gauges and the corresponding N hours of actual rainfall data as the training data set to train the GCN model. The node features of the rain gauges are used as model input, and the corresponding actual rainfall data is used as model output. Then, use the trained GCN model to identify abnormal rain gauges, eliminate the identified abnormal stations, and record the number of rain gauges operating normally as K.

[0009] Step 3: Determine candidate locations for additional rain gauge stations: Use Kriging interpolation method to spatially interpolate the actual rainfall data of K rain gauge stations hourly to generate rainfall field data P for the target basin. 站点插值 , then for the J grids divided in step 1, each grid center point has two sets of rainfall data, namely radar rainfall data P and 雷达 and interpolated rainfall field data P 站点插值 , calculate P at each grid center hourly 雷达 and P 站点插值 Relative error RE j , j=1,2,…,J,, statistics of RE in each period j >N number of grids in D2 D2,t , set N D2,t The period with a rainfall distribution difference of >D3 belongs to the period with two obvious differences in rainfall distribution. The total number of such periods is T. The thresholds D2 and D3 are determined according to the natural geographical conditions and data of the basin. The relative error RE of each grid is calculated in T periods. j The average value of γ grid points is arranged from large to small, and the first γ grid center points are taken as candidate locations for the replacement station, numbered 1, 2, ..., γ;

[0010] Step 4, accuracy evaluation of different rainfall station network optimization schemes: Set 1 to γ ​​additional rainfall stations and generate a total of γ additional station schemes. The γ additional station schemes are to add a station at candidate position number 1, to add stations at candidate positions numbered 1 and 2, and so on, to add stations at candidate positions numbered 1 and 2, ..., γ. The total number of rainfall stations corresponding to the γ additional station schemes is K+1, K+2, ..., K+γ respectively.

[0011] A machine learning model is used to predict rainfall conditions for different rain gauge replacement schemes. For each rain gauge location s, a feature vector X(s) is constructed, including the geographical location, terrain characteristics, meteorological characteristics of the station, and the radar rainfall data corresponding to the grid. The feature vectors and actual rainfall data of all K rain gauges are then used as training data sets to train a machine learning model f, with the feature vector X of the rain gauge as the model input and the corresponding actual rainfall data as the model output. The trained model is then used to predict rainfall conditions for each candidate location s* based on its feature vector X(s * ), predicting rainfall at its location for:

[0012]

[0013] For all T time periods, the rainfall data of each candidate location of the γ replacement station schemes are predicted period by period, and the prediction results under each replacement station scheme are spatially interpolated using the Kriging interpolation method to generate the corresponding rainfall field data;

[0014] For T periods, calculate RE again period by period j >The number of grids in D2 is expressed as That is, RE in the tth period j > the number of grids in D2, t=1,2,…,T, and then calculate the average percentage reduction of the number of grids in all time periods per:

[0015]

[0016] The larger the per value is, the closer the basin rainfall data after the station is added is to the radar rainfall data.

[0017] Step 5: Determine the optimal layout of the rain gauge network: Considering the construction cost, maintenance cost, and potential socioeconomic benefits of the rain gauge network, a penalty function ω is constructed to represent the adverse effects of field conditions on the construction of rain gauges. Let C represent the construction cost, W represent the maintenance cost, and B represent the socioeconomic benefits. C, W, and B are all functions of the additional station plan and change with the plan. The penalty function ω is defined as:

[0018]

[0019] Where: α, β, is the weight coefficient;

[0020] Combining per and ω, we construct the optimization objective function φ so that it is positively correlated with per and negatively correlated with ω, that is:

[0021] φ=μ1×per-μ2×ω (6)

[0022] Where: μ1, μ2 are weight coefficients;

[0023] For each additional station scheme, the corresponding φ value is calculated, and the scheme with the highest φ value is selected as the optimal rainfall station layout scheme.

[0024] Furthermore, the same grid division of the target area according to the grid distribution of the radar rainfall data described in step 1 is specifically as follows: parsing the radar rainfall data file, extracting its spatial resolution, and determining the basic unit of the grid; then based on the spatial position of the radar data, the grids are systematically numbered using the row and column coordinate method, with the row numbers increasing from north to south and the column numbers increasing from west to east.

[0025] Furthermore, the node features of the rain gauges constructed in step 2 are specifically as follows: the geographical location relationship between the rain gauges is converted into a graph structure, each rain gauge is regarded as a node in the graph structure, the edges between the nodes represent the spatial correlation between the rain gauges, and the weights of the edges are set by distance measurement or similarity measurement based on geographical distance; the node features of the rain gauges include rainfall time series features and geographic information features; the rainfall time series features include rainfall statistical features and time feature information in historical periods; the geographic information features include the latitude and longitude, altitude, and terrain features of each rain gauge.

[0026] Furthermore, the identification of abnormal rainfall stations using the trained GCN model described in step 2 is specifically as follows: for each rainfall station, for all N time periods, the reconstruction error RE predicted by the trained GCN model is calculated:

[0027]

[0028] Where: and are the measured value at the rainfall station at time n and the predicted value based on the GCN model;

[0029] Rainfall stations with RE>D1 are marked as abnormal stations, where the threshold D1 is determined in combination with regional weather conditions and natural geographical characteristics.

[0030] Furthermore, the terrain characteristics in step 4 include altitude, slope, and slope direction; and the meteorological characteristics include temperature, humidity, and wind speed.

[0031] The beneficial effect of the present invention is that the method of the present invention combines radar rainfall measurement data to achieve effective planning and optimization of rainfall station networks under complex terrain conditions in mountainous areas, thereby improving the accuracy and reliability of precipitation observations in mountainous watersheds, and providing more accurate data support for the management and prevention of flood disasters in mountainous areas.

[0032] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0034] The present invention discloses a method for optimizing the layout of rainfall station network in mountainous area based on radar rainfall measurement data. Figure 1 As shown, the method includes the following steps:

[0035] Step 1: Collect rainfall data and divide the target area into grids.

[0036] For the target area where the layout of the rainfall station network is to be optimized, first collect the actual rainfall data P of all M rainfall stations in the area. 雨量站 and P 雨量站 Radar rainfall data corresponding to the time series P 雷达 , giving priority to rainfall during the flood season. Secondly, the target area is divided into identical grids (J in total) based on the grid distribution of the radar rainfall data, and each grid is assigned a unique identifier. Specifically, the radar rainfall data file is parsed and its spatial resolution is extracted to determine the basic unit of the grid. Then, based on the spatial location of the radar data, the grids are systematically numbered using a row and column coordinate method, with row numbers increasing from north to south and column numbers increasing from west to east. This ensures that each grid has a unique identifier, facilitating subsequent data processing and analysis.

[0037] Step 2: Identify abnormal rainfall stations.

[0038] Considering that existing rain gauges may have operational anomalies, a GCN (graph convolutional network) model is used to identify abnormal sites to ensure that the existing station network consists of normally operating sites. First, the node features of the rain gauges are constructed. If the rainfall data of each rain gauge is N hours, the node features of M rain gauges and the corresponding N hours of actual rainfall data are used as the training data set to train the GCN model. The node features of the rain gauges are used as model input, and the corresponding actual rainfall data is used as model output. The trained GCN model is then used to identify abnormal rain gauges. The identified abnormal sites are removed, and the number of normally operating rain gauges is recorded as K. Subsequent steps are performed only based on K rain gauges, and abnormal sites are no longer considered.

[0039] The specific process is as follows:

[0040] 1) Construct node features of rain gauges. The geographical location relationship between rain gauges is converted into a graph structure. Each rain gauge is regarded as a node in the graph structure. The edges between nodes represent the spatial correlation between rain gauges. The weight of the edges is set by a distance metric (such as Euclidean distance) or a similarity metric based on geographic distance. The node features of rain gauges include rainfall time series features and geographic information features. The rainfall time series features include rainfall statistical features of historical periods (such as the mean and variance of precipitation in the past few hours or days), time features (year, month, season, time of day, etc.); geographic information features include the latitude and longitude, altitude, and terrain features of each rain gauge.

[0041] 2) Initialize the model parameters and train the GCN model using the training dataset. Given the model layers L, the number of neurons in each layer d1, and the activation function σ, the lth (l = 1, 2, ..., L) layer of the GCN model can be expressed as:

[0042]

[0043] Where: H (l) is the node feature matrix of the lth layer; is the adjacency matrix after adding the self-loop; is the degree matrix; W (l) is the weight matrix of the lth layer; σ is the activation function, such as ReLU.

[0044] 3) Abnormal station identification. For each rainfall station, for all N time periods, calculate the reconstruction error RE after the trained GCN model prediction:

[0045]

[0046] Where: and are the measured values ​​at the rainfall station at time n and the predicted values ​​based on the GCN model, respectively.

[0047] Rainfall stations with RE>D1 are marked as abnormal stations, where the threshold D1 is determined in combination with regional weather conditions and natural geographical characteristics.

[0048] Step 3: Determine the candidate locations for additional rain gauge stations.

[0049] The Kriging interpolation method is used to spatially interpolate the rainfall data of K rain gauges hourly to generate the rainfall field data P of the target basin. 站点插值 So far, for the J grids drawn in step 1, each grid center point has two sets of rainfall data, namely radar rainfall data P and 雷达 and interpolated rainfall field data P 站点插值 Calculate P at each grid center hourly 雷达 and P站点插值 The relative error RE of the two rainfall j (j=1,2,…,J), statistics of RE in each period j >N number of grids in D2 D2,t , set N D2,t The period with >D3 belongs to the period with obvious difference in rainfall distribution. The total number of such periods is T, where the thresholds D2 and D3 are determined according to the natural geographical conditions of the basin and data conditions. Then, the relative error RE of each grid is calculated in T periods. j The average value of , arranged from large to small, takes the first γ grid center points as candidate locations for filling stations, numbered 1, 2, …, γ respectively.

[0050] Step 4: Evaluate the accuracy of different rainfall station network optimization schemes.

[0051] Set up 1 to γ ​​additional rain gauges and generate a total of γ additional station plans. The γ additional station plans are to add a station at candidate position number 1, to add stations at candidate positions numbered 1 and 2, and so on, to add stations at candidate positions numbered 1 and 2, …, γ. The total number of rain gauges corresponding to the γ additional station plans is K+1, K+2, …, K+γ respectively.

[0052] Use machine learning models (such as random forests, gradient boosting trees, etc.) to predict rainfall conditions for different rain gauge replacement schemes. For each rain gauge location s, construct a feature vector X(s), including the geographical location of the station, terrain characteristics (including altitude, slope, slope direction, etc.), meteorological characteristics (including temperature, humidity, wind speed, etc.) and radar rainfall data corresponding to the grid. Then use the feature vectors and actual rainfall data of all K rain gauges as training data sets to train the machine learning model f, with the feature vector X of the rain gauge as model input and the corresponding actual rainfall data as model output. Then use the trained model to predict rainfall conditions for each candidate location s* based on its feature vector X(s * ), predicting rainfall at its location (i.e., the rainfall measured if there is a surface rain gauge here) is:

[0053]

[0054] For all T time periods, the rainfall data of each candidate location of the γ replacement station schemes are predicted period by period, and the prediction results under each replacement station scheme are spatially interpolated using the Kriging interpolation method to generate the corresponding rainfall field data;

[0055] For T periods, calculate RE again period by period j >The number of grids in D2 is expressed as (RE in period t j> the number of grids in D2, t = 1, 2, ..., T), and then calculate the average percentage reduction of the number of grids in all time periods per:

[0056]

[0057] The larger the per value is, the closer the basin rainfall field data after the supplementary station is to the radar rainfall data; the more significantly the ability of the ground rain gauge network to capture the basin rainfall conditions is improved.

[0058] Step 5: Determine the optimal layout plan of the rainfall station network.

[0059] Based on actual economic efficiency, a station network layout optimization objective function is constructed to determine the optimal rainfall station network layout plan. Considering the construction cost, maintenance costs, and potential socioeconomic benefits of the rain gauge station, a penalty function ω is constructed to represent the adverse impact of field conditions on the construction of rain gauges. Let C represent the construction cost, W represent the maintenance cost, and B represent the socioeconomic benefits. C, W, and B are all functions of the station replenishment plan and change with the plan. The penalty function ω can be defined as:

[0060]

[0061] Where: α, β, is the weight coefficient.

[0062] Combining per and ω, we construct the optimization objective function φ so that it is positively correlated with per and negatively correlated with ω, that is:

[0063] φ=μ1×per-μ2×ω (6)

[0064] Where: μ1 and μ2 are weight coefficients, which are allocated according to importance.

[0065] For each additional station scheme, the corresponding φ value was calculated, and the scheme with the highest φ value was selected as the optimal rain gauge layout scheme. This scheme not only maximizes the accuracy of rainfall observations, but also ensures the economic rationality of construction and maintenance, achieving a comprehensive technical and economic optimization.

[0066] Finally, it should be noted that the above description is only used to illustrate the technical solution of the present invention and is not intended to limit it. Although the present invention has been described in detail with reference to the preferred arrangement scheme, those skilled in the art should understand that the technical solution of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for optimizing the layout of rainfall station networks in mountainous areas based on radar rainfall data, characterized in that: The method comprises the following steps: Step 1: Collect rainfall data and divide the target area into grids: Collect the actual rainfall data P of all M rainfall stations in the target area. 雨量站 and P 雨量站 Radar rainfall data corresponding to the time series P 雷达 , divide the target area into the same grids according to the grid distribution of radar rainfall data, record a total of J, and assign a unique identifier to each grid; Step 2: Identify abnormal rain gauges: Use the graph convolutional network model (GCN) to identify abnormal rain gauges. First, construct the node features of the rain gauges. If the actual rainfall data of each rain gauge is N hours, then use the node features of M rain gauges and the corresponding N hours of actual rainfall data as the training data set to train the GCN model. The node features of the rain gauges are used as model input, and the corresponding actual rainfall data is used as model output. Then, use the trained GCN model to identify abnormal rain gauges, eliminate the identified abnormal stations, and record the number of rain gauges operating normally as K. Step 3: Determine candidate locations for additional rain gauge stations: Use Kriging interpolation method to spatially interpolate the actual rainfall data of K rain gauge stations hourly to generate rainfall field data P for the target basin. 站点插值 , then for the J grids divided in step 1, each grid center point has two sets of rainfall data, namely radar rainfall data P and 雷达 and interpolated rainfall field data P 站点插值 , calculate P at each grid center hourly 雷达 and P 站点插值 Relative error RE j , j=1,2,…,J, count RE in each period j >N number of grids in D2 D2,t , set N D2,t The period with a rainfall distribution difference of >D3 belongs to the period with two obvious differences in rainfall distribution. The total number of such periods is T. The thresholds D2 and D3 are determined according to the natural geographical conditions and data of the basin. The relative error RE of each grid is calculated in T periods. j The average value of the grid points is arranged from large to small, and the first γ grid center points are taken as candidate locations for the replacement station, numbered 1, 2, ..., γ; Step 4, accuracy evaluation of different rainfall station network optimization schemes: Set 1 to γ ​​additional rainfall stations and generate a total of γ additional station schemes. The γ additional station schemes are to add a station at candidate position number 1, to add stations at candidate positions numbered 1 and 2, and so on, to add stations at candidate positions numbered 1 and 2, ..., γ. The total number of rainfall stations corresponding to the γ additional station schemes is K+1, K+2, ..., K+γ respectively. A machine learning model is used to predict rainfall conditions for different rain gauge replacement schemes. For each rain gauge location s, a feature vector X(s) is constructed, including the geographical location, terrain characteristics, meteorological characteristics of the station, and the radar rainfall data corresponding to the grid. The feature vectors and actual rainfall data of all K rain gauges are then used as training data sets to train a machine learning model f, with the feature vector X of the rain gauge as the model input and the corresponding actual rainfall data as the model output. The trained model is then used to predict rainfall conditions for each candidate location s* based on its feature vector X(s * ), predicting rainfall at its location for: For all T time periods, the rainfall data of each candidate location of the γ replacement station schemes are predicted period by period, and the prediction results under each replacement station scheme are spatially interpolated using the Kriging interpolation method to generate the corresponding rainfall field data; For T periods, calculate RE again period by period j >The number of grids in D2 is expressed as That is, RE in the tth period j > the number of grids in D2, t=1,2,…,T, and then calculate the average percentage reduction of the number of grids in all time periods per: The larger the per value is, the closer the basin rainfall data after the station is added is to the radar rainfall data. Step 5: Determine the optimal layout of the rain gauge network: Considering the construction cost, maintenance cost, and potential socioeconomic benefits of the rain gauge network, a penalty function ω is constructed to represent the adverse effects of field conditions on the construction of rain gauges. Let C represent the construction cost, W represent the maintenance cost, and B represent the socioeconomic benefits. C, W, and B are all functions of the additional station plan and change with the plan. The penalty function ω is defined as: Where: α, β, is the weight coefficient; Combining per and ω, we construct the optimization objective function φ so that it is positively correlated with per and negatively correlated with ω, that is: φ=μ1×per-μ2×ω (6) Where: μ1, μ2 are weight coefficients; For each additional station scheme, the corresponding φ value is calculated, and the scheme with the highest φ value is selected as the optimal rainfall station layout scheme.

2. The method for optimizing the layout of a rainfall station network in a mountainous area based on radar rainfall data according to claim 1, characterized in that: The specific steps of dividing the target area into the same grid according to the grid distribution of radar rainfall data in step 1 are as follows: parsing the radar rainfall data file, extracting its spatial resolution, and determining the basic unit of the grid; then, based on the spatial position of the radar data, using the row and column coordinate method to systematically number the grids, with row numbers increasing from north to south and column numbers increasing from west to east.

3. The method for optimizing the layout of a rainfall station network in a mountainous area based on radar rainfall data according to claim 1, characterized in that: The node features of the rain gauges constructed in step 2 are specifically as follows: the geographical location relationship between the rain gauges is converted into a graph structure, each rain gauge is regarded as a node in the graph structure, the edges between the nodes represent the spatial correlation between the rain gauges, and the weights of the edges are set by distance measurement or similarity measurement based on geographical distance; the node features of the rain gauges include rainfall time series features and geographic information features; rainfall time series features include rainfall statistical features and time feature information in historical periods; geographic information features include the latitude and longitude, altitude, and terrain features of each rain gauge.

4. The method for optimizing the layout of a rainfall station network in a mountainous area based on radar rainfall data according to claim 1, characterized in that: The specific method for identifying abnormal rainfall stations using the trained GCN model in step 2 is as follows: for each rainfall station, for all N time periods, calculate the reconstruction error RE predicted by the trained GCN model: Where: and are the measured value at the rainfall station at time n and the predicted value based on the GCN model; Rainfall stations with RE>D1 are marked as abnormal stations, where the threshold D1 is determined in combination with regional weather conditions and natural geographical characteristics.

5. The method for optimizing the layout of a rainfall station network in a mountainous area based on radar rainfall data according to claim 1, characterized in that: The terrain characteristics in step 4 include altitude, slope, and slope direction; the meteorological characteristics include temperature, humidity, and wind speed.

Images