An integrated method for analyzing and predicting spatiotemporal laws of water quality based on large-scale basins

Abstract

The present invention relates to the technical field of watershed pollution prediction, and in particular to an integrated method for analyzing and predicting the spatiotemporal laws of water quality based on a large scale in a watershed, including S1, collecting different water quality data from monitoring stations in each section, and using random forest to fill in missing water quality data; S2, using variance analysis to analyze the spatiotemporal differences of water quality indicators; S3, using a combination of a comprehensive pollution index method and a dynamic degree analysis method to analyze the temporal and spatial variation laws of water quality; S4, using an ARIMA model to predict the changing trends of lake and river pollution factors. The present invention uses variance analysis to perform a difference analysis on water quality indicators, and uses Tukey HSD to test the significance of the difference; the present invention uses a comprehensive water quality pollution index method and a dynamic degree analysis method to analyze the spatiotemporal laws of the water environment in the watershed from the time and space dimensions, combining qualitative and quantitative analysis; and uses ARIMA to predict the changing trends of the main pollutants in the river for a period of time in the future.

Description

Technical Field

[0001] The present invention relates to the technical field of watershed pollution prediction, and in particular to an integrated method for analyzing and predicting spatiotemporal laws of large-scale water quality in a watershed. Background Art

[0002] The Taoge River Basin is the main area of ​​the Taoge Taishui system in the western part of the Taihu Lake Basin, and is also an important catchment area in the upper reaches of the Taihu Lake Basin. The improvement of the water ecological environment quality in the Taoge River Basin is extremely important for improving the water quality of the main rivers entering Taihu Lake and improving the quality and benefits of the water ecological environment of Taihu Lake. However, due to the continuous development of urbanization in recent years, shallow lakes in the Taoge River Basin have become eutrophic and blue algae have frequently erupted. Therefore, how to take effective environmental management measures to improve the water environment of shallow lakes has become an urgent problem to be solved.

[0003] At present, the water quality evaluation methods used in rivers and lakes include single factor evaluation method, fuzzy comprehensive evaluation method, grey system theory, principal component analysis method, etc. At present, the research on water environment in river basins mostly takes a single lake or river as the research object, comprehensively evaluates the changes in water quality; or combines multivariate statistical methods to analyze the spatiotemporal changes in water environment quality in a certain area. At present, there are more studies on water quality in the Taihu Lake Basin, and fewer spatiotemporal studies and analyses on the water quality of multiple rivers in the Taoge Basin, and combined with deep learning models to predict the water quality trend of rivers over a long period of time in the future. Summary of the invention

[0004] The technical problem solved by the present invention is as follows: using the comprehensive pollution index and dynamic degree to explore the water quality change characteristics in time and space, scientifically analyzing the temporal and spatial change characteristics of the water quality of lakes and rivers in the basin, and deeply analyzing the shortcomings of the ecological environment faced by the basin; and combining the moving difference average autoregression model to predict the future change trends of the main pollution factors of the rivers entering and leaving the lakes in the basin, deeply analyzing the causes of the change trends, and proposing feasible scientific suggestions.

[0005] The technical solution adopted by the present invention is: an integrated method for analyzing and predicting the spatiotemporal laws of water quality in a large-scale basin, comprising the following steps:

[0006] S1. Preprocess the water quality data of each section monitoring station of the main rivers entering and leaving the lake in the basin: including collecting different water quality data of each section monitoring station, using random forest to fill in the missing water quality data, and normalizing them;

[0007] S2. Use variance analysis to analyze the temporal and spatial differences of water quality indicators of the main rivers entering and leaving the lake in the basin, and use Tukey HSD to test the significance of the differences;

[0008] S3. Combine the comprehensive pollution index method with the dynamic degree analysis method to analyze the temporal and spatial variation patterns of the water quality of the main rivers entering and leaving the lake in the basin, and divide the main rivers entering and leaving the lake in the basin into different water systems based on the lake-river connectivity, and then analyze the degree of influence of different water systems on water quality;

[0009] S4. Use the Autoregressive Moving Average (ARIMA) model to predict the changing trends of the main pollution factors of lakes and rivers in the future, and obtain the range of changes in the pollution factors, providing technical support for the management of river basin water environment.

[0010] Furthermore, the step S1 specifically includes:

[0011] S11. Use the numpy and pandas libraries in Python to batch collect water quality data of the main inlet and outlet river sections of the Taoge River Basin and save them in an Excel spreadsheet;

[0012] S12. Use the random forest algorithm to complete the missing water quality data of lake and river sections;

[0013] S13. Normalize the completed water quality data so that data of different water quality factor types are normalized to the same order of magnitude.

[0014] Furthermore, the step S2 specifically includes:

[0015] S21. From the two dimensions of time and space, variance analysis was used to analyze the difference of water quality of the main rivers entering and leaving the lake in the basin, and it was determined that the water quality factors have temporal and spatial differences;

[0016] S22, assuming that at the significance level If the F calculated by ANOVA is significant, then Tukey HSD is calculated based on the difference in sample means. The calculated value is called HSD, which is as follows: Among them, q value and significance level The number of mean values ​​in the sample, k, is related to the error degrees of freedom, nk, and n j is sample type j, MSE is the root mean square error, which refers to the expected value of the square of the difference between the parameter estimate and the true value of the parameter; the difference between the means of any pair of samples exceeds the HSD value, indicating that the difference between the pair of means is significant; note that the statistic HSD requires that the size of all samples is equal.

[0017] Furthermore, the step S3 specifically includes:

[0018] S31. Comprehensively evaluate the water quality of the main inflow and outflow lakes in the basin using the comprehensive pollution index method, and qualitatively analyze the changes in the water quality of the basin on a time scale according to the overall dynamic characteristics of the comprehensive pollution index of lake and river water quality and the monthly dynamics;

[0019] S32. There are many river systems in the basin, which are widely distributed. We divide the river systems according to the structure of the river basin, divide the different river systems according to the connectivity between lakes and rivers, and analyze the spatial variation characteristics of water quality in different water systems.

[0020] Furthermore, the step S4 specifically includes:

[0021] S41, ADF and KPSS unit root methods were used to test the stationarity of water quality data, and Akaike information criterion (AIC) and Bayesian information criterion (BIC) were used to determine the order of ARIMA model;

[0022] S42. Use the ARIMA model after order determination to fit the water quality time series, use the maximum likelihood estimation method to estimate the ARIMA model parameters, and test the autocorrelation of the ARIMA model fitting residuals;

[0023] S43. After the residual test, the ARIMA model with a fixed order is used to predict the changing trend of the main pollution factors of lakes and rivers in the future, and the range value of the main pollution factors in the future is obtained.

[0024] The beneficial effects of the present invention are:

[0025] 1. First, variance analysis was used to analyze the differences in water quality indicators of the main rivers entering and leaving the lakes in the basin, and Tukey HSD was used to test the significance of the differences; secondly, the comprehensive water quality pollution index method and dynamic degree analysis method were used to analyze the temporal and spatial laws of the water environment in the basin from the two dimensions of time and space, combining qualitative and quantitative methods. According to the connectivity between lakes and rivers, the basin was divided into different water systems, and the degree of influence of different water systems on water quality was analyzed; finally, the ARIMA model was used to predict the changing trend of the main pollutants in the river in the future.

[0026] 2. The temporal and spatial variation patterns of lake and river water quality within a large scale are analyzed from the basin level, and the analysis and prediction are integrated into a set of methods, which can provide technical guidance for the temporal and spatial analysis and prediction of water quality within a large basin. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a flow chart of the integrated method for analyzing and predicting large-scale spatiotemporal laws of a watershed according to the present invention;

[0028] Figure 2 The distribution of monitoring stations of the main inflow and outflow rivers in the Taoge River Basin in the present invention;

[0029] Figure 3 It is the concentration value of pollution factors such as total phosphorus, total nitrogen, and permanganate index in the Taoge River Basin in the present invention;

[0030] Figure 4 This is a 30-day trend chart of the main pollution factors of the Zhongzhong River in the Taoge River Basin of the present invention;

[0031] Figure 5 This is a 30-day changing trend chart of the main pollution factors of Xilicaohe River in Taoge River Basin in the present invention. DETAILED DESCRIPTION

[0032] The present invention is further described below in conjunction with the accompanying drawings and embodiments. This figure is a simplified schematic diagram, which only illustrates the basic structure of the present invention in a schematic manner, and therefore it only shows the components related to the present invention.

[0033] like Figure 1 As shown in the figure, an integrated method for analyzing and predicting the spatiotemporal laws of water quality based on a large scale in a river basin has the following steps:

[0034] S1. Preprocess the water quality data of each section monitoring station of the main rivers entering and leaving the lake in the Taoge River Basin: including collecting different water quality data of each section monitoring station, using random forest to fill in the missing value data of water quality, and normalizing it;

[0035] S2. ANOVA was used to analyze the temporal and spatial differences of water quality indicators of the main rivers entering and leaving the lake in the Taoge River Basin, and Tukey HSD was used to test the significance of the differences;

[0036] S3. Combine the comprehensive pollution index method with the dynamic degree analysis method to analyze the temporal and spatial variation patterns of the water quality of the main rivers entering and leaving the lakes in the Taoge River Basin. Based on the lake-river connectivity, the main rivers entering and leaving the lakes in the Taoge River Basin are divided into different water systems, and then the degree of influence of different water systems on water quality is analyzed.

[0037] S4. Use the Autoregressive Imaging Model (ARIMA) to predict the changing trends of the main pollution factors of lakes and rivers in the next 30 days, and obtain the range of changes in the pollution factors, providing technical support for the water environment management in the Taoge River Basin.

[0038] The specific experimental methods are as follows:

[0039] 1. Study area and water quality index selection of Taoge River Basin

[0040] Taoge River Basin (119 ° 48 ′ ~120 ° 01 ′ N, 31 ° 37′ ~31 ° 86 ′ E) has a total area of ​​1707.52 km2. It is a typical shallow grass lake basin in southern Jiangsu. It is located in the southern part of the lower reaches of the Yangtze River and has a northern subtropical monsoon climate. Figure 2 As shown in the figure, dots represent sites for basin monitoring. The basin is adjacent to Taihu Lake in the east, Changdang Lake in the west, Yixing in the south, and the Yangtze River in the north. The main rivers entering and leaving the lake include Xiaxi River, Huangli River, Zhonggan River, Wunan River, Caoqiao River, etc. The Taoge Basin includes Gehu Lake and Changdang Lake. Among them, Gehu Lake, as the main lake in the upper reaches of Taihu Lake and the middle and lower reaches of the Yangtze River, is responsible for connecting the two major water bodies of the Yangtze River and Taihu Lake, regulating water volume, preventing floods and waterlogging, and transporting clean water. Similarly, Gehu Lake is the hub of the Yangtze River to Taihu Xinmeng River Project, which sends Yangtze River water to Taihu Lake through the Taige Canal.

[0041] The intensified eutrophication of water quality and the excessive total phosphorus have always been the key issues in the Taoge Basin. The eutrophication of river and lake water bodies is inseparable from nutrients such as nitrogen and phosphorus, and will cause the dissolved oxygen in the water body to reach an oversaturated state. This study selected 14 water quality monitoring sections in 7 rivers in the Taoge Basin. Most of the monitoring sections are national and provincial river sections. The water quality data comes from the high-frequency monitoring of the section water quality automatic monitoring station, and the monitoring frequency is 4h. When studying and analyzing the spatiotemporal water quality of rivers and lakes, the water quality data are sorted into monthly averages. When predicting the future trend of changes in river and lake water quality pollutants, the water quality data are sorted into daily averages. Obtaining high-frequency monitoring water quality data can more accurately grasp the changes in water quality in the Taoge Basin. For 26 national and provincial river sections, 21 water quality indicators are based on Class III water limits. It was found that 8 water quality indicators such as DO, CODmn, TP, and NH3-N exceeded the standard more frequently. When analyzing the water quality of typical shallow lakes in the middle and lower reaches of the Yangtze River (Taihu Lake, Chaohu Lake, Poyang Lake, etc.), most of them use DO, pH, CODmn, TP, TN, chl-a, NH3-N and other indicators as water quality evaluation factors. In summary, this study selected six indicators including dissolved oxygen (DO), pH, permanganate index (CODmn), total phosphorus (TP), total nitrogen (TN), and ammonia nitrogen (NH3-N) as the main indicators for the analysis and evaluation of water eutrophication in the Taoge Basin.

[0042] 2. Analysis of temporal and spatial differences in water quality

[0043] The variance analysis was used to analyze the differences of the water quality indicators of the main rivers entering and leaving the lake in the Taoge River Basin, and the Tukey HSD test was used to test the significance of the differences. The results of the time and space difference test of the water quality indicators of the monitoring sections of the rivers entering and leaving the lake in the Taoge River Basin are shown in Tables 1 and 2. The water quality indicators of the monitoring sections showed significant differences in time (p<0.05); in space, the p values ​​of total phosphorus and pH were 0.110 and 0.935, respectively. These two water quality indicators had high similarity in space, especially pH, because the amplitude of pH change was relatively small and the differences among the monitoring sections were not obvious. However, dissolved oxygen, permanganate index, total nitrogen and ammonia nitrogen showed significant differences in space (p<0.05); in order to more accurately describe the degree of temporal and spatial changes in water quality in the Taoge River Basin, it is also necessary to conduct comprehensive pollution index analysis and dynamic analysis on the water quality of the main rivers entering and leaving the lake in this area.

[0044] Table 1 Test of water quality time dimension differences

[0045]

[0046]

[0047] Table 2 Test of water quality spatial dimension differences

[0048]

[0049] 3. Analysis of spatiotemporal variation characteristics of water quality of major rivers flowing into and out of the Taoge River Basin

[0050] In the analysis of the temporal and spatial variation of water quality in the Taoge River Basin, the comprehensive pollution index analysis method and dynamic degree analysis are mainly used; the comprehensive pollution index method is used to comprehensively evaluate the river water quality, and the calculation formula is as follows:

[0051]

[0052]

[0053] Where P i is the pollution index of water quality factor i; C i is the measured concentration of water quality factor i, in mg / L; n is the number of water quality factors; P is the comprehensive pollution index.

[0054] The dynamic degree analysis method was used to analyze the change degree of the comprehensive pollution index of the Taoge River Basin during the study period. The calculation formula is:

[0055]

[0056] Where S is the dynamic degree of the comprehensive pollution index of the river during the study period, P mTo study the comprehensive pollution index of river m in the final year and month, P n It is the comprehensive pollution index of river n in the initial year of the study.

[0057] 3.1 Analysis of temporal variation characteristics of water quality in the basin

[0058] In July 2020, the average comprehensive pollution index of the main rivers entering and leaving the lake in the Taoge River Basin was 0.901, and in October 2021 it was 0.812, a decrease of 9.88% compared with July 2020. The overall water quality of the main rivers entering and leaving the lake has improved to a certain extent. The comprehensive pollutant index of the main rivers entering and leaving the lake in the Taoge River Basin from July 2020 to October 2021 is shown in Table 3. The water quality of the main rivers entering and leaving the lake in the Taoge River Basin was the worst in January 2021, and the water quality was the best in October 2021. The average comprehensive pollution index of the main rivers entering and leaving the lake in the dry season (2020.12-2021.3) was 1.062, and the average in the flood season (2021.6-2021.10) was 0.861. The flood season decreased by 18.93% compared with the dry season, indicating that the overall water quality in the Taoge River Basin in the flood season has improved compared with the dry season.

[0059] According to the monthly dynamics and overall dynamics of the comprehensive pollution index of river water quality from October 2020 to October 2021, the water quality trends of the main rivers entering and leaving the lake in the Taoge Basin can be divided into two categories. One category is that the overall dynamics of the comprehensive pollution index of rivers from October 2020 to October 2021 is negative, and the monthly dynamics are alternating between positive and negative, indicating that the overall river water quality is improving, but the water quality has a certain fluctuation trend and cannot be continuously improved. The other category is that the overall dynamics of the comprehensive pollution index of rivers from October 2020 to October 2021 is positive, and the monthly dynamics are alternating between positive and negative, indicating that the water quality is fluctuating and deteriorating as a whole.

[0060] The water quality change trends of the main rivers entering and leaving the lake in the Taoge Basin from 2020 to 2021 are shown in Table 4. The total dynamics of Xiaxi River, Zhonggan River and Xilicao River are negative, and the water quality is generally improving, among which the water quality of Xilicao River has improved most significantly; the total dynamics of Huangli River, Caoqiao River, Wunan River and Wuyiyun River are positive, and the water quality is generally fluctuating and deteriorating, among which the water quality change of Caoqiao River is the most severe. In general, the overall improvement of rivers with improved water quality is greater than that of rivers with poor water quality, indicating that the water environment governance in the Taoge Basin has achieved certain results.

[0061] Table 3 Comprehensive pollutant index of major rivers flowing into and out of the Taoge River Basin in 2020-2021

[0062]

[0063] Table 4 Trends in water quality of major rivers flowing into and out of the Taoge River Basin from 2020 to 2021

[0064]

[0065] 3.2 Analysis of spatial variation characteristics of water quality in the basin

[0066] The Taoge River Basin has numerous and widely distributed water systems. The division based on the basin water system structure is adopted to analyze the spatial change characteristics of water quality of the main rivers in and out of the lakes in the Taoge River Basin from 2020 to 2021.

[0067] According to the change of the comprehensive pollution index of the main rivers in and out of the Taoge River Basin over time, the changes in the comprehensive pollution index are graded, as shown in Table 5. From October 2020 to mid-October 2021, the comprehensive pollution index of Xiaxi River, Zhonggan River and Xili Caohe River was on a downward trend, the dynamic degree of the comprehensive pollution index of Xiaxi River decreased significantly, and the water quality condition was significantly improved; the dynamic degree of Zhonggan River decreased significantly, and the dynamic degree of the comprehensive pollution index of Xili Caohe River decreased significantly overall, and the water quality condition improved significantly. In October 2021, the comprehensive pollution index of Huangli River, Wunan River, Wuyi Canal and Caoqiao River was on an upward trend compared with October 2020. Among them, the dynamic degree of the comprehensive pollution index of Huangli River and Wunan River increased slightly, by 6.19% and 6.31% respectively, and the dynamic degree of Caoqiao River increased significantly, by 20.75%, and the water quality condition was relatively severe. As of October 2021, the main rivers entering and leaving the lake in the Taoge River Basin had the highest comprehensive pollution index in the Xilicao River, and the lowest in the Zhonggan River.

[0068] Table 5 Changes in pollution index of each river

[0069]

[0070] Based on the connectivity between rivers and lakes, the main rivers in and out of the Taoge Basin are divided into three categories: the Changdang Lake-Gehu Lake system, whose main rivers include Xiaxi River, Huangli River and Zhonggan River; the Gehu Lake-West Taihu Lake system, whose main rivers include Wunan River and Caoqiao River; and the Beijing-Hangzhou Grand Canal-Xinmeng River system, including the Wuyi Grand Canal and Xili Caohe River. The above three types of water systems are connected with the water systems along the river, forming a network water system that meets the east and the west and intersects the north and the south.

[0071] It can be seen from Table 6 that the comprehensive pollutant index of the Changdang Lake-Gehu Lake water system was 0.690 in October 2021, a decrease of 7.26% compared with 0.744 in October 2020. The dynamic degree of the comprehensive pollution index decreased slightly, and the water quality improved; although the comprehensive pollution index of the Beijing-Hangzhou Canal-Xinmeng River water system was the highest, the decline was the largest, by 9.99%, and the dynamic degree of the comprehensive pollution index was greatly reduced, and the water quality improved significantly. Due to the opening of the Xinmeng River, the Yangtze River water with better water quality was introduced and the river channel along the line was widened, and the water quality of the tributaries and main streams along the line gradually improved; the comprehensive pollution index of the Gehu Lake-West Taihu Lake water system in October 2021 increased by 13.14% compared with October 2020, the dynamic degree of the comprehensive pollution index increased significantly, and the water quality was poor.

[0072] Table 6 Dynamic changes in the comprehensive pollution index of each water system from 2020 to 2021

[0073]

[0074] 4. Analysis of temporal variation characteristics of main pollution factors in lake and river water quality in the basin

[0075] The above qualitative analysis of the changes in water quality in the Taoge River Basin on a time scale shows that although the water quality of the main rivers flowing into and out of the Taoge River Basin has improved from 2020 to 2021, problems such as increased eutrophication and excessive total phosphorus are still key issues in the basin; Figure 3 As shown in the data, from July 2020 to October 2021, the average mass concentration of total phosphorus in the main rivers entering and leaving the lake in the Taoge Basin showed a downward trend. The large fluctuation of water concentration between months is the basic characteristic of TP in the Taoge Basin, from 0.151 mg.L in July 2020 to 0.151 mg.L in October 2021. -1 Down to 0.129mg.L in October 2021 -1 , down 14.48%, which reflects the influence of changes in human activities such as water pollution control in the basin. However, as of October 2021, only the total phosphorus concentration in Zhonggan River can reach the Class IV evaluation standard, and the total phosphorus concentrations in other rivers are all at the Class V evaluation standard; the average mass concentration of total nitrogen has dropped from 2.318 mg.L in July 2020 to 2.318 mg.L in July 2020. -1 Down to 2.101mg.L in October 2021 -1, a decrease of 9.36%. The average concentration of total nitrogen in August 2021 was the lowest during this period. Overall, the total nitrogen index in the Taoge River Basin during the flood season (June-September) is better than that in the non-flood season. This may be because the dilution effect of large water volume is obvious in areas dominated by point source pollution, and the dilution effect during the flood season is the main influencing factor. Similar to total phosphorus, the total nitrogen water quality in Zhonggan River is the best and can meet the III evaluation standard. From July 2020 to October 2021, the average mass concentration of the permanganate index of the main rivers in and out of the lake in the Taoge River Basin fluctuated significantly. Except for January and July 2021, the average quality had a significant upward trend. The average quality in the rest of the period was decreasing, from 4.539 mg.L in July 2020. -1 Down to 3.868mg.L in October 2021 -1 , a decrease of more than 14.78%. In general, the concentrations of major pollutants such as total phosphorus, total nitrogen and permanganate index in the main rivers entering and leaving the lake in the Taoge River Basin showed an overall downward trend from 2020 to 2021. However, among the many pollutant factors, phosphorus has always been the key factor affecting river water quality. Therefore, the key to improving the water quality of the Taoge River Basin lies in the control of phosphorus.

[0076] 5. Predict the future trend of major pollution factors in the Taoge River Basin based on the ARIMA model

[0077] 5.1 Construction of ARIMA Model

[0078] The ARIMA model is a commonly used non-stationary time series prediction model and a generalized model of the autoregressive moving average (ARMA). It combines the autoregressive (AR) process and the moving average (MA) process to establish a composite model of the time series. The ARMA model can only analyze stationary time series. t , ARIMA transforms Y into t Transform into a stationary time series X t It should be noted that if the number of differences is too large, there will be a certain loss of data information. The steps of construction are as follows:

[0079] 1. Use ADF and KPSS unit root methods to test the stability of water pollution factor time series data;

[0080] 2. If it does not pass the stationarity test, perform d-order difference on the time series data to convert it into a stationary time series;

[0081] 3. Use Akaike information criterion and Bayesian information criterion to determine the order of the model;

[0082] 4. Use the determined-order model to fit the time series and use the maximum likelihood estimation method to estimate the model parameters;

[0083] 5. Test the autocorrelation of the model fitting residuals. The closer the residuals are to the white noise distribution, the more sufficient the useful information in the time series is extracted.

[0084] 6. After passing the residual test, the specified ARIMA model is used for time series forecasting; if the residual test fails, the model will be re-ordered.

[0085] 5.2 Model prediction results display

[0086] After testing the stability of sample data and determining the model order, this study used the corresponding ARIMA model to predict the total phosphorus, permanganate index and total nitrogen sample data of Zhonggan River and Xilicao River for the next 30 days. The prediction results are as follows: Figure 4 and Figure 5 As shown. Taking the permanganate index of Zhonggan River as an example, based on the data from July 1, 2020 to September 30, 2021 as the training set, the permanganate index from October 1, 2021 to October 30, 2021 is predicted. It can be seen from the figure that the degree of fit between the predicted value and the observed value is relatively consistent. The gray shaded part in the figure is the interval range of the predicted value of the ARIMA model, indicating that the predicted value of the future water quality factor varies within this interval, which to a certain extent reflects the future change trend of the river water quality factors. From October 1, 2021 to October 30, 2021, the predicted intervals of total phosphorus, permanganate index and total nitrogen in Zhonggan River and Xilicao River are shown in Table 7. As shown in Table 7, the range of total phosphorus, permanganate content and total nitrogen in Zhonggan River during this period was 0.068-0.189, 3.74-6.49 and 0.349-2.69 respectively, and the range of water quality change in Xilicao River was 0.048-0.310, 1.59-4.91 and 5.13-9.38 respectively, all in mg.L-1. Combined with the above analysis of the water quality of Zhonggan River and Xilicao River, the forecast interval trend is consistent with the current water quality of the two rivers, but it is worth noting that the total nitrogen concentration of Xilicao River will change greatly in the next month. In summary, the ARIMA model can accurately obtain the trend of river water quality changes for a long period of time in the future.

[0087] Table 7 Range of water quality factors of Zhonggan River and Xilicao River from October 1 to October 30, 2021

[0088]

[0089] Based on the above ideal embodiments of the present invention, the relevant staff can make various changes and modifications without departing from the technical concept of the present invention through the above description. The technical scope of the present invention is not limited to the contents of the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. An integrated method for analyzing and predicting the spatiotemporal laws of water quality in a large-scale basin, characterized by: The steps are as follows: S1. Preprocess the water quality data of each monitoring station of the rivers entering and leaving the lake in the basin: including collecting different water quality data of each monitoring station, using random forest to fill in the missing water quality data, and performing normalization processing; S2. Use variance analysis to analyze the temporal and spatial differences of water quality indicators of rivers entering and leaving the lake, and use Tukey HSD to test the significance of the differences; The step S2 specifically includes: S21. From the two dimensions of time and space, variance analysis is used to analyze the difference of lake and river water quality, and the temporal and spatial differences of water quality factors are identified; S22, assuming that at the significance level If the F calculated by ANOVA is significant, then Tukey HSD is calculated based on the difference between sample means. The Tukey HSD formula is as follows: Among them, q value and significance level The number of mean values ​​in the sample, k, is related to the error degrees of freedom, nk, n j is sample type j, MSE is the root mean square error, which refers to the expected value of the square of the difference between the parameter estimate and the true value of the parameter; when the difference between the mean values ​​of any pair of samples exceeds the HSD value, it indicates that the difference between the pair of mean values ​​is significant; S3. Combine the comprehensive pollution index method with the dynamic degree analysis method to analyze the temporal and spatial variation patterns of water quality of rivers entering and leaving the lake in the basin, and divide the rivers entering and leaving the lake into different water systems based on the lake-river connectivity, and then analyze the degree of influence of different water systems on water quality; S4. Use the differential moving average autoregressive model to predict the changing trend of the main pollution factors of lakes and rivers in the future and obtain the range of pollution factor changes; The step S4 specifically includes: S41, ADF and KPSS unit root methods were used to test the stationarity of water quality data, and Akaike information criterion and Bayesian information criterion were used to determine the order of ARIMA model; S42. Use the ARIMA model after order determination to fit the water quality time series, use the maximum likelihood estimation method to estimate the ARIMA model parameters, and test the autocorrelation of the ARIMA model fitting residuals; S43. After the residual test, the ARIMA model with a fixed order is used to predict the changing trend of the main pollution factors of lakes and rivers in the future, and the range value of the main pollution factors in the future is obtained.

2. The integrated method for analyzing and predicting the spatiotemporal regularity of water quality in a large-scale basin according to claim 1, characterized in that: The step S1 specifically includes: S11. Use numpy and pandas libraries to collect water quality data of lake and river sections in batches; S12. Use the random forest algorithm to complete the missing water quality data of lake and river sections; S13. Normalize the completed water quality data so that data of different water quality factor types are normalized to the same order of magnitude.

3. The integrated method for analyzing and predicting the spatiotemporal regularity of water quality based on a large scale of a river basin as claimed in claim 1, characterized in that: The step S3 specifically includes: S31. Comprehensively evaluate the water quality of lakes and rivers using the comprehensive pollution index method, and qualitatively analyze the changes in water quality on a time scale based on the overall dynamic characteristics of the comprehensive pollution index and monthly dynamics of lake and river water quality; S32. Based on the division of river basin water system structure, different water systems are divided according to the connectivity between lakes and rivers, and the spatial variation characteristics of water quality in different water systems are analyzed.

Images