A method for estimating the chemical oxygen demand (COD) storage of reservoirs over a long time series
By combining remote sensing data and time series analysis with the dam construction year and reservoir capacity model, the problem of estimating reservoir COD reserves was solved, and accurate calculation of reservoir and national total reservoir reserves was achieved.
Patent Information
- Application Number
- CN202310399068.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-04-13
- Publication Date
- 2026-01-30
- Estimated Expiration
- 2043-04-13
AI Technical Summary
Existing technologies are insufficient to effectively estimate the chemical oxygen demand (COD) reserves in reservoirs, especially in reservoir areas where measured data is lacking and there are high spatiotemporal variability, making it impossible to accurately calculate the total COD reserves of reservoirs nationwide.
The reservoir water surface was identified by combining remote sensing data with normalized water index and threshold. The construction year of the dam was determined by Census X-11 time series decomposition. A correlation model between the average COD concentration of the reservoir and the construction year of the dam was established. The COD storage of the reservoir was calculated by combining the total reservoir capacity.
It has achieved accurate estimation of COD storage in reservoirs, especially for small-area reservoirs, and further estimated the total COD storage of reservoirs nationwide through upscaling method, thus solving the problem of spatiotemporal variation in reservoir areas.
Smart Images
Figure CN116644553B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of optical satellite remote sensing, water environment assessment, and carbon cycle. It mainly uses historical archived reservoir chemical oxygen demand (COD) data and satellite remote sensing images to remotely estimate reservoir water storage and extract the dam construction year. Then, it constructs an estimation model for reservoir COD concentration and finally combines statistical data to achieve long-term estimation of total reservoir COD storage. Background Technology
[0002] COD is an important indicator of organic pollution in water bodies and a crucial indicator for monitoring surface water environmental quality. COD is also closely related to dissolved organic carbon (DOC), the largest organic carbon pool in water bodies; changes in COD storage directly reflect changes in carbon storage. However, current estimations of reservoir COD or DOC storage are largely nonexistent. Although some recent studies have attempted to estimate lake DOC storage using field surveys or water color remote sensing data, these measurement-based estimation methods cannot be applied to reservoir areas with insufficient data. Low-spatial-resolution water color remote sensing data is also difficult to apply to reservoir areas with small water areas. Furthermore, the highly spatiotemporally variable distribution and water volume of reservoirs increase the difficulty of estimating COD or DOC storage in reservoirs nationwide. While it is difficult to directly monitor reservoir COD storage using water color remote sensing data, water color satellite remote sensing data has been successfully applied to water area extraction and lake water storage change estimation. Related methods have been applied to lake groups in different regions of my country, especially the Qinghai-Tibet Plateau lake group, where water area and water storage have shown high variability in recent years. Therefore, if a model can be built to estimate the COD concentration in reservoirs, the COD storage of reservoirs can be calculated by further combining the estimated water storage with water color remote sensing satellite data; for the total COD storage of reservoirs nationwide, an upscaling estimate can be made by comprehensively utilizing annual statistical data. Summary of the Invention
[0003] The purpose of this invention is to provide a method for estimating the chemical oxygen demand (COD) reserves of a reservoir over a long time period.
[0004] To achieve the above technical objectives, the present invention adopts the following technical solution:
[0005] A method for estimating the chemical oxygen demand (COD) storage of a long-term reservoir includes:
[0006] Reservoir water surface identification is achieved by combining remote sensing data with normalized water index and threshold.
[0007] The Census X-11 time series decomposition was performed on the identified reservoir surface area to obtain the contribution ratio of trend components. The Mann-Kendall test was performed on the trend components to determine the year corresponding to the abrupt change point in the reservoir area, which was used as the year of dam construction.
[0008] Obtain the average COD concentration of the reservoir in the time series monitoring, and establish a fitting model of the correlation between the COD concentration and the construction year of the dam.
[0009] The average COD concentration of the reservoir in different years is estimated by using the correlation fitting model combined with the year, and the product of the average COD concentration of the reservoir with the total reservoir capacity is calculated, that is, the COD storage of the reservoir in different years.
[0010] As a preferred embodiment, the method for identifying the reservoir water surface is as follows:
[0011] Threshold segmentation is used to initially obtain the water coverage area;
[0012] Using the reservoir dataset as seed points, water cell dilation is performed. The intersection of the dilation result and the water coverage area determined by the OTSU algorithm is taken as the water area coverage range of different reservoirs.
[0013] As a preferred implementation method, the OTSU algorithm is used for binary segmentation to initially obtain the water body coverage area.
[0014] As a preferred embodiment, the remote sensing data is selected from Landsat series satellite time series data.
[0015] As a preferred implementation method, a quadratic equation is established using the average COD concentration of the reservoir monitored over time as the dependent variable and the year of dam construction as the independent variable, which serves as the correlation fitting model.
[0016] In a preferred embodiment, the total reservoir capacity is determined as follows: the waterline is obtained based on the identified reservoir water surface, and the water surface elevation is determined by overlaying and analyzing it with DEM data. The total reservoir capacity is then calculated based on the water surface elevation and DEM data. Furthermore, ArcMap software is used to convert the identified reservoir water surface into a line to obtain the waterline.
[0017] As a preferred embodiment, the total reservoir capacity is calculated based on the following formula:
[0018] (3)
[0019] In the formula, V is the total storage capacity; A i D is the area of pixel i, calculated based on the spatial resolution of the remote sensing image; i H represents the water depth of pixel i; i The water surface elevation for pixel i; DEMi is the ground elevation of pixel i before the dam was built; m is the number of pixels in the reservoir.
[0020] In a preferred embodiment, the method further includes establishing a correlation model between the total waterway capacity and the total reservoir capacity identified by remote sensing, and then using the correlation model to calculate the total COD storage of the reservoir in different years by upscaling. Further, the correlation model is in exponential form.
[0021] The present invention has the following beneficial effects:
[0022] (1) Remote sensing of typical reservoir water storage and dam construction year. Water storage is an important indicator for calculating reservoir COD storage, and reservoir capacity exhibits significant spatiotemporal variation characteristics with dam construction. This invention proposes a reservoir capacity estimation method by utilizing satellite remote sensing data with spatiotemporal coverage advantages, reservoir datasets, and digital elevation models (DEMs), while obtaining the reservoir construction year based on time-series water area data obtained from remote sensing.
[0023] (2) Another important indicator for calculating reservoir COD storage is the COD concentration in the water body, and the COD concentration in reservoirs also shows obvious spatiotemporal variation patterns. This invention combines the dam construction year and existing COD monitoring data to construct a rapid estimation model for the annual average COD concentration of reservoirs.
[0024] (3) Remote sensing images can only capture images of reservoirs with relatively large resolvable areas, while the number of small reservoirs and their COD storage also need to be given special attention. Based on the estimated COD storage and total storage capacity statistics of typical reservoirs, this invention constructs a method for upscaling the estimation of the total COD storage in the country. Attached Figure Description
[0025] Figure 1 The relationship between reservoir COD concentration and the year the dam was built.
[0026] Figure 2 This is a time series of COD and water storage in Chinese reservoirs from 1950 to 2020. Detailed Implementation
[0027] Taking reservoirs in China as an example, this invention details a specific implementation method for estimating the chemical oxygen demand (COD) storage of long-term reservoirs, as follows:
[0028] (1) Atmospheric correction of remote sensing data. Time series data of Landsat series satellites with high spatial resolution were downloaded from the U.S. Geological Survey (USGS), and then atmospheric correction of remote sensing data was performed based on the Quick Atmospheric Correction (QUAC) tool of the ENVI software platform to obtain the remote sensing reflectance of ground objects.
[0029] (2) Identification of typical reservoir water surfaces. The Normalized Difference Water Index (NDWI) is calculated from the remote sensing reflectance obtained by atmospheric correction; the NDWI calculation formula is shown in Equation (1), where R Green and R NIR These are the remote sensing reflectance values for the green light and near-infrared bands, respectively.
[0030] (1)
[0031] After calculating the NDWI, the OTSU algorithm is further used for binary segmentation to determine the water coverage area. Finally, using the China Reservoir Dataset (CRD) obtained from Zonodo as seed points, water cell dilatation is identified to determine the water area coverage of different reservoirs.
[0032] (3) Analysis of the construction year of the reservoir dam. For data with records, the construction year of the dam is directly assigned. For other reservoirs with unknown dam construction years, Census X-11 time series decomposition is performed after obtaining the water area to obtain the contribution ratio of seasonal components, trend components, and irregular components. Then, the Mann-Kendall test is performed on the trend component to determine the dam construction year, that is, the year corresponding to the abrupt change point of the reservoir area. The Census X-11 method is defined as shown in Equation (2), where S, T, and I correspond to the nodular component, trend component, and irregular component, respectively.
[0033] (2)
[0034] (4) Estimation of water storage in typical reservoirs. In ArcMap software, the obtained reservoir water surface is converted into a line to obtain the water edge line. The water edge line is then overlaid with the DEM to determine the reservoir water surface elevation (H). The total reservoir capacity is then calculated from the water surface elevation and the DEM. The calculation formula is as follows (3):
[0035] (3)
[0036] In the formula, V is the total storage capacity; A i D is the area of pixel i, calculated based on the spatial resolution of the remote sensing image; i H represents the water depth of pixel i; iThe water surface elevation for pixel i; DEM i is the ground elevation of pixel i before the dam was built; m is the number of pixels in the reservoir.
[0037] (5) Construction of reservoir COD concentration model. By analyzing the time series monitoring COD concentration of 47 lakes in 2021 and the year of dam construction, it was found that there is a significant quadratic equation (Equation (4)) between the annual average COD concentration of lakes and the year of dam construction, that is: the COD concentration of reservoirs gradually decreases in the first 30 years after the dam is built, and then the COD concentration of reservoirs increases. Figure 1 ).
[0038] (4)
[0039] (6) Estimation of COD storage in typical reservoirs. Combining the dam construction year analyzed by remote sensing and formula (3), the average COD concentration (C) of reservoirs in different years can be obtained. COD ), and then multiplied by the water storage (V) obtained through step (4). water It can calculate the COD capacity (V) of a typical reservoir. COD The calculation formula is as shown in equation (5):
[0040] (5)
[0041] (7) Upscaling estimation of total COD storage capacity in China. Through the above steps (1)-(6), the total storage capacity of typical reservoirs in China in different years since 1950 can be estimated. By comparing the total storage capacity statistics of China since 1979, it is found that the total storage capacity of typical reservoirs (Vtypical) and the total storage capacity of China (Vtotal) show a significant exponential correlation, as shown in equation (6). Therefore, based on equation (6), the storage capacity of reservoirs built in different years since 1950 can be calculated, and combined with equation (4), the total COD storage capacity of Chinese reservoirs since 1950 can be estimated on an upscaling basis. During the period from 1950 to 2020, the storage capacity and COD storage capacity of Chinese reservoirs increased by 3 times and 1.67 times, respectively, and the growth rate has been particularly significant since 2000. Figure 2 ).
[0042] (6).
Claims
1. A method for estimating long-term chemical oxygen demand storage of a reservoir, characterized by, The method comprises the following steps: reservoir water surface is identified by remote sensing data combined with normalized water index and threshold value; CensusX-11 time series decomposition is performed on the identified reservoir water surface area to obtain the contribution ratio of the trend component, and Mann-Kendall test is performed on the trend component to determine the year corresponding to the abrupt change point of the reservoir area as the dam construction year; time series monitoring of the average COD concentration of the reservoir is performed, and a correlation fitting model with the dam construction year is established; the correlation fitting model is used to calculate the average COD concentration of the reservoir in different years, and the product of the total reservoir capacity is calculated, that is, the COD storage of the reservoir in different years.
2. The method of claim 1, wherein, The method for identifying the reservoir water surface is as follows: atmospheric correction is performed on the remote sensing data to obtain remote sensing reflectivity data, and normalized water index is calculated to preliminarily obtain the water covered area by threshold segmentation; water pixel inflation is performed with the reservoir data set as the seed point, and the intersection of the inflation result and the water covered area determined by OTSU algorithm is taken as the water area coverage range of different reservoirs.
3. The method of claim 2, wherein, OTSU algorithm is used for binary segmentation to preliminarily obtain the water covered area.
4. The method according to claim 1 or 2, characterized in that, The remote sensing data is selected from Landsat series satellite time series data.
5. The method of claim 1, wherein, A monomial quadratic equation is established with the time series monitoring of the average COD concentration of the reservoir as the dependent variable and the dam construction year as the independent variable as the correlation fitting model.
6. The method of claim 1, wherein, The method for determining the total reservoir capacity is as follows: the water edge line is obtained based on the identified reservoir water surface, and is superimposed and analyzed with DEM data to determine the water surface elevation of the reservoir, and the total reservoir capacity is calculated based on the water surface elevation and DEM data.
7. The method of claim 6, wherein, ArcMap software is used to convert the identified reservoir water surface into a line to obtain the water edge line.
8. The method according to claim 1 or 6, characterized in that, The total reservoir capacity is calculated based on the following formula: In the formula, V is the total storage capacity; A i is the area of pixel i, calculated according to the spatial resolution of remote sensing image; D i is the water depth of pixel i; H i is the water surface elevation of pixel i; DEM i is the ground elevation of pixel i before the construction of the dam; and m is the number of reservoir pixels.
9. The method of claim 1 or 6, wherein, Furthermore, a correlation model of the total reservoir capacity of the reservoir identified by remote sensing and the total reservoir capacity is established, and the correlation fitting model is used to scale up the calculation of the total COD storage of the reservoir in different years.
10. The method of claim 9, wherein, The correlation model is in the form of an exponential function.