A rainfall interpolation method considering precipitation duration

Abstract

The present invention discloses a rainfall interpolation method considering precipitation duration, and belongs to the field of data processing technology. The present invention overcomes the problem that the existing traditional precipitation time interpolation method often ignores the change of precipitation with duration, resulting in the accuracy of the interpolation result being limited by introducing the precipitation time interpolation method of precipitation duration. Starting from the perspective of considering the effect of time on precipitation interpolation, the present invention analyzes and improves the original precipitation interpolation method, considers the influence of precipitation duration on precipitation, and proposes a time interpolation method considering precipitation duration, which is divided into five aspects, namely, determining the research area and research period, initialization and precipitation trend judgment, climate state data reorganization, construction of hourly precipitation distribution ratio, and construction of hourly precipitation data. The present invention will provide more accurate precipitation data support for fields such as climate research, agricultural production, and water resources management, and help to better understand and respond to the impact of climate change on human society and ecosystems.

Description

Technical Field

[0001] The invention belongs to the technical field of data processing, and in particular relates to a rainfall interpolation method considering precipitation duration. Background Art

[0002] Precipitation interpolation refers to the use of known precipitation data to estimate the precipitation conditions in unknown areas. The existing precipitation interpolation methods are mainly divided into the following categories:

[0003] Linear interpolation: Assume that precipitation is evenly distributed throughout the day. However, this method does not take into account the actual occurrence and duration of precipitation events, so it is not accurate enough.

[0004] Random generation method: Use probability distribution (such as Poisson distribution or exponential distribution) to randomly generate hourly precipitation based on the frequency and intensity of precipitation in the observed data. This method can simulate the randomness and unevenness of precipitation, but requires sufficient historical data to build a probabilistic model.

[0005] Climate statistical method: Analyze historical climate data, determine the probability and intensity distribution of precipitation in a specific time period, and then distribute daily precipitation data to each hour. This method takes into account the seasonal and daily variation characteristics of precipitation, but the calculation complexity is relatively high.

[0006] Model-driven approach: Use the output of high-resolution weather forecast models or climate models to estimate hourly precipitation through dynamical downscaling. This approach can provide detailed information about the physical mechanisms of precipitation processes, but requires expensive computing resources and professional technical support.

[0007] Artificial neural network and machine learning methods: Use deep learning algorithms (such as recurrent neural networks) to predict hourly precipitation. These algorithms can learn precipitation patterns from a large amount of historical data. This method can simulate complex precipitation processes very accurately, but requires a large amount of training data and high-performance computing facilities. Summary of the invention

[0008] In view of the problems in the prior art, the present invention provides a rainfall interpolation method taking precipitation duration into consideration, aiming to solve the limitations of the existing precipitation data time interpolation methods.

[0009] The technical solution adopted by the present invention is as follows:

[0010] A rainfall interpolation method considering precipitation duration includes the following steps:

[0011] S1: Collect daily precipitation data and three-hourly precipitation data in the study area;

[0012] S2: Identify the daily precipitation trend based on the daily precipitation data and three-hourly precipitation data;

[0013] S3: Read the precipitation data of each geographical location at each time point by grid point, combine the precipitation data of the regional climate state, calculate the similar precipitation day with the same precipitation trend as the calculation grid point and the precipitation difference less than the set value, find the three-hour precipitation corresponding to the calculation grid point in the three-hour precipitation data of the similar precipitation day, construct the similar precipitation day sequence of the calculation grid point, calculate the proportion of the corresponding three-hour precipitation to the total precipitation of the day, and integrate to form the climate state three-hour data group;

[0014] S4: average the three-hourly climate state data set to obtain the three-hourly average precipitation ratio of the climate state;

[0015] S5: Multiply the obtained climatological three-hourly precipitation ratio by the daily precipitation to obtain the reconstructed climatological three-hourly precipitation data.

[0016] Preferably, in order to record the precipitation trend at each time point, S2 initializes a string array with the same dimension as the precipitation data, and determines the precipitation trend by comparing the precipitation on the day before and after the calculation day. According to the comparison result, the precipitation trend is marked as increasing, decreasing or stable.

[0017] The specific approach is to assume a 7-day precipitation sequence (such as [10,12,9,14,14,8,10]), and create a string array with the same dimension to mark the precipitation trend. By comparing the daily precipitation with the previous day, if it increases, it is marked as "increasing", if it decreases, it is marked as "decreasing", and if it is the same, it is marked as "stable". The final generated array is such as ["","Increasing","Decreasing","Increasing","Stable","Decreasing","Increasing"], which records the daily precipitation change trend.

[0018] Preferably, the daily precipitation data and three-hourly precipitation data of the study area collected in S1 include near-ground air temperature, near-ground air pressure, near-ground air specific humidity, near-ground total wind speed, downward short-wave radiation from the ground, downward long-wave radiation from the ground, and ground precipitation rate. The time resolution of the collected data is 3 hours, and the horizontal spatial resolution is 0.1°.

[0019] Preferably, the collected data comes from the China regional ground meteorological element driven data set.

[0020] Preferably, days with precipitation differences less than or equal to 1% in S3 are similar precipitation days.

[0021] Preferably, when more than ten similar corresponding three-hourly precipitation amounts are found in S3, the proportion of the three-hourly precipitation amount of each similar precipitation day to the precipitation amount of the similar precipitation day is calculated.

[0022] Ten samples are needed to ensure the significance of the samples. The core of this study is to select samples with significance as a data set. The selection criteria for similar rainfall are strict, so the number of similar rainfall days selected is not large. The selection of 10 similar rainfall days can ensure the significance of the data.

[0023] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0024] 1. Compared with traditional methods, the present invention has significantly improved the accuracy of precipitation prediction. By considering the impact of precipitation duration on precipitation, the present invention can more accurately predict the precipitation trend in different regions and seasons, thereby improving the accuracy and reliability of precipitation prediction.

[0025] 2. The present invention also has strong versatility and scalability. It can be applied to precipitation interpolation in different regions and seasons, and provides a simple and effective precipitation data processing tool for the fields of meteorology, environmental science, agricultural production, etc.

[0026] 3. The present invention not only improves the accuracy and reliability of temperature forecasting, but also expands the application field of precipitation data, providing more reliable and accurate precipitation data support for research and application in related fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is a flow chart of the present invention;

[0028] Figure 2 It is a schematic diagram of precipitation trend in the present invention;

[0029] Figure 3 This is a graph of precipitation results in 2018 before and after interpolation in the present invention. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of this application.

[0031] like Figure 1 As shown in FIG. 1 , a rainfall interpolation method considering the duration of precipitation is described. The precipitation data of CMFD in 2018 in the study area (90-110°E, 28-40°N) is used to illustrate the method of the present invention as follows:

[0032] S1: Download precipitation data from 1979 to 2018 from the China Regional Surface Meteorological Element Driven Dataset to obtain daily precipitation data and three-hourly precipitation data in the study area;

[0033] The daily precipitation data and three-hourly precipitation data collected in the study area include near-surface air temperature, near-surface air pressure, near-surface air specific humidity, near-surface total wind speed, downward short-wave radiation from the ground, downward long-wave radiation from the ground, and ground precipitation rate. The time resolution of the collected data is 3 hours, and the horizontal spatial resolution is 0.1°.

[0034] S2: Identify the daily precipitation change trend based on the daily precipitation data and the three-hour precipitation data; in order to record the precipitation trend (increasing, decreasing or stable) at each time point, initialize a string array with the same dimension as the precipitation data. Determine the precipitation change trend by comparing the precipitation on the day before and after the calculation day. Based on the comparison results, mark the precipitation change trend as "increasing" (the precipitation on the previous day is less than the precipitation on the next day), "decreasing" (the precipitation on the previous day is greater than the precipitation on the next day) or "stable" (the precipitation on the previous day is equal to the precipitation on the next day). Figure 2 As shown, the daily precipitation data from June 19, 2018 to June 26, 2018, 28.15°E, 94.95°N. June 21-22, 2018 is defined as the increasing period, and June 21-22, 2018 is defined as the decreasing period.

[0035] S3: Read precipitation data at each moment grid by grid, combine with precipitation data of regional climate state, calculate similar precipitation days with the same precipitation trend of the climate state of the calculation grid point and with precipitation difference less than or equal to 1%, find the three-hour precipitation corresponding to the calculation grid point in the three-hour precipitation data of the similar precipitation day, construct a similar precipitation day sequence of the calculation grid point, and start calculating the proportion of the corresponding three-hour precipitation to the total precipitation of the day when more than ten similar precipitations are found, and integrate to form a climate state three-hour data group;

[0036] S4: average the three-hourly climate state data set to obtain the three-hourly average precipitation ratio of the climate state;

[0037] S5: Multiply the obtained climatological three-hour precipitation ratio by the daily precipitation to obtain the reconstructed climatological three-hour precipitation data. So far, the calculation result of a grid point is obtained. Read all longitudes and latitudes in a loop and repeat the above operation to obtain the calculation results of each grid point. The precipitation results for 2018 before and after interpolation are as follows: Figure 3 As shown, the correlation coefficient between the reconstructed 2018 precipitation and the observed value reaches 0.65, indicating that the present invention has a high accuracy in reproducing the three-hourly precipitation trend.

[0038] The above-mentioned embodiments only express the specific implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the protection scope of the present application. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the technical solution concept of the present application, and these all belong to the protection scope of the present application.

Claims

1. A rainfall interpolation method considering precipitation duration, characterized in that: The following steps are involved: S1: Collect daily precipitation data and three-hour precipitation data in the study area, including near-surface temperature, near-surface air pressure, near-surface air specific humidity, near-surface total wind speed, ground-down shortwave radiation, ground-down longwave radiation, and ground precipitation rate; S2: Identify the daily precipitation trend based on the daily precipitation data and three-hourly precipitation data; S3: Read the precipitation data of each geographical location at each time point by grid point, combine the precipitation data of the regional climate state, calculate the similar precipitation day with the same precipitation trend as the grid point and the precipitation difference less than the set value, find the three-hour precipitation corresponding to the grid point in the three-hour precipitation data of the similar precipitation day, construct the similar precipitation day sequence of the grid point, calculate the proportion of the corresponding three-hour precipitation to the total precipitation of the day, and integrate to form the climate state three-hour data group; S4: average the three-hourly climate state data set to obtain the three-hourly average precipitation ratio of the climate state; S5: Multiply the obtained climatological three-hourly precipitation ratio by the daily precipitation to obtain the reconstructed climatological three-hourly precipitation data.

2. A rainfall interpolation method considering precipitation duration according to claim 1, characterized in that: In order to record the precipitation trend at each time point, S2 initializes a string array with the same dimension as the precipitation data. The precipitation trend is determined by comparing the precipitation before and after the calculation day. According to the comparison result, the precipitation trend is marked as increasing, decreasing or stable.

3. The rainfall interpolation method considering precipitation duration according to claim 1, characterized in that: The daily precipitation data and three-hourly precipitation data of the study area collected in S1 have a temporal resolution of 3 h and a horizontal spatial resolution of 0.1°.

4. A rainfall interpolation method considering precipitation duration according to claim 3, characterized in that: The collected data comes from the China regional ground meteorological element driven dataset.

5. A rainfall interpolation method considering precipitation duration according to any one of claims 1 to 4, characterized in that: In S3, days with precipitation differences less than or equal to 1% are considered similar precipitation days.

6. A rainfall interpolation method considering precipitation duration according to any one of claims 1 to 4, characterized in that: When more than ten similar corresponding three-hourly precipitation amounts are found in S3, the proportion of the three-hourly precipitation amount of each similar precipitation day to the precipitation amount of the similar precipitation day is calculated.

Images