Method for observing temporal and spatial variation rule of cyanobacterial bloom in Taihu Lake based on Sentinel-2 remote sensing image

Abstract

The invention relates to the technical field of environment monitoring, in particular to a method for observing the temporal and spatial change rule of cyanobacterial bloom in the Taihu Lake based on a Sentinel-2 remote sensing image. The method comprises the following steps: S1, obtaining remote sensing data: obtaining a remote sensing image of Sentinel-2 and multispectral remote sensing data corresponding to the remote sensing image; s2, contour extraction of a water area: calculating to obtain a Taihu Lake water area range mask according to the multispectral remote sensing data in the step S1; s3, division of a Taihu Lake water area observation area: dividing the Taihu Lake water area contour obtained in the step S2 into seven areas; s4, extracting the contour of the cyanobacterial bloom: calculating the intensity and range of the cyanobacterial bloom according to the multispectral remote sensing data in the step S1; and S5, observing the temporal and spatial change rule of cyanobacterial bloom: analyzing the spatial change characteristics of the cyanobacterial bloom in the Taihu Lake according to the result obtained in the step S4. By analyzing the remote sensing image of the existing Sentinel-2 and the multispectral remote sensing data corresponding to the remote sensing image, the change rule of the cyanobacterial bloom in the Taihu Lake can be conveniently summarized and evaluated.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of environmental monitoring, in particular to a method for observing spatiotemporal variation law of cyanobacterial bloom in Taihu Lake based on Sentinel-2 remote sensing image. BACKGROUND

[0002] Inland lakes are an important source of global freshwater resources. Water bloom refers to the phenomenon of overproduction of certain algae caused by excessive accumulation of nutrients such as nitrogen and phosphorus in freshwater bodies, which is one of the important characteristics of water eutrophication. Among them, cyanobacterial bloom is the most common and serious eutrophication problem in inland lakes. Water bloom will lead to a decrease in water biodiversity, disrupt the balance of aquatic ecosystems, and threaten water health, causing serious economic losses to society. Therefore, timely, comprehensive, and accurate monitoring and understanding of water bloom occurrence is crucial for water bloom prevention and control. Traditional water eutrophication monitoring methods mainly rely on manual sampling to estimate the number of algae in the laboratory environment to determine the degree of water eutrophication. Although this method can achieve high monitoring accuracy, it is time-consuming and labor-intensive, and is easily affected by weather and hydrological conditions. Therefore, it is only suitable for small water bodies and difficult to continuously observe in space and time. Compared with traditional water bloom monitoring methods, remote sensing monitoring technology has the advantages of wide detection range, low cost, real-time dynamic, and can effectively make up for the limitations of traditional manual monitoring methods, achieving long-term and large-scale dynamic monitoring of cyanobacterial bloom. Therefore, it is widely used in lake cyanobacterial bloom monitoring.

[0003] At present, most of the researches on cyanobacterial bloom monitoring at home and abroad focus on the selection of remote sensing image data and the extraction and inversion of cyanobacterial bloom. The research area is mainly large water area rivers and lakes, and the research method is mainly to select remote sensing index to extract water bloom and quantitatively invert parameters such as chlorophyll-a content and algal density. Sentinel-2 is a high-resolution multispectral imaging satellite developed by the European Space Agency (ESA), with two satellites, 2A and 2B. Sentinel-2 multispectral image contains 13 spectral bands, with a width of about 290 kilometers, and the highest spatial resolution of 10 m. The complementary revisit period of the two satellites is 5 days, and only 3 days in high latitude areas. In general, Sentinel-2 has the advantages of high spatial resolution, short playback period, rich spectral bands, open source acquisition, etc., so it has been widely used in cyanobacterial bloom monitoring. Manuel et al. (2021) evaluated the feasibility of using Sentinel-2 satellite to identify cyanobacterial bloom and chlorophyll-a, and found that the correlation of green band was the best, and the Toming index had the best correlation effect among the used indexes, indicating that Sentinel-2 could become a powerful tool for cyanobacterial bloom monitoring research; Gernez et al. (2023) used Sentinel-2 satellite data and field monitoring data to build a red tide database containing 27 phytoplankton species, indicating that Sentinel-2 satellite can effectively detect high-concentration harmful algal blooms, providing a new perspective for red tide remote sensing monitoring. The basic information of Sentinel-2 multispectral bands is shown in Table 1.

[0004]

[0005] Lake Taihu is one of the lakes with serious cyanobacterial bloom in China. Cyanobacterial bloom can be seen throughout the year since the late 1980s. From 1990 to 2007, the water quality of Lake Taihu declined rapidly, and cyanobacterial bloom occurred frequently and severely. Although the management started in the mid-to-late 1990s, the effect was not obvious. In late May 2007, a large-scale cyanobacterial bloom occurred in Lake Taihu. Since then, the management of cyanobacterial bloom in Lake Taihu has been fully launched. Since 2007, the cumulative investment of all levels of finance and society has exceeded 300 billion yuan for the management of Lake Taihu, and significant results have been achieved. As of 2024, Lake Taihu has completed ecological dredging of 63.5 million cubic meters, built 18 ecological safety buffer zones, and effectively controlled the outbreak of cyanobacterial bloom. In recent years, under the effect of management measures, the intensity of cyanobacterial bloom in Lake Taihu has decreased, but it still needs to be continuously monitored and managed.

[0006] The present study selects Sentinel-2 remote sensing images in Taihu area from 2021 to 2023, extracts cyanobacterial bloom information by using the NDVI method, and verifies it by using ground observation data, obtains the distribution range, area, concentration and other information of cyanobacterial bloom in a large area of water in the south of Taihu Lake, and tracks its dynamic changes, including the growth, diffusion, migration path and recession process of the bloom, so as to provide comprehensive and accurate data support for studying the formation mechanism and development law of cyanobacterial bloom. SUMMARY

[0007] The purpose of the present application is to provide a method for observing the spatiotemporal variation law of cyanobacterial bloom in Taihu Lake based on Sentinel-2 remote sensing images, so as to provide reliable technical support for the dynamic observation and variation law grasping of cyanobacterial bloom in Taihu Lake.

[0008] In order to achieve the above-mentioned purpose, the technical scheme of the present application is as follows: The method for observing the spatiotemporal variation law of cyanobacterial bloom in Taihu Lake based on Sentinel-2 remote sensing images comprises the following steps: S1, obtaining remote sensing data: obtaining Sentinel-2 remote sensing images and multispectral remote sensing data corresponding to the remote sensing images; S2, contour extraction of water area: calculating the NDWI of each group of data according to the multispectral remote sensing data in step S1, and completely excluding the land part according to the threshold value to obtain a mask of the water area range of Taihu Lake for application in subsequent cropping, wherein, or ; NDWI is the normalized water index, R Green is the green band reflectance, R NIR is the near-infrared band reflectance, R Band3 is the Sentinel-2 3rd band reflectance, R Band5 is the Sentinel-2 5th band reflectance; S3, division of the observation area of Taihu Lake water area: dividing the Taihu Lake water area contour obtained in step S2 into 7 regions, including a central lake center and 6 edge regions around the lake center; S4, contour extraction of cyanobacterial bloom: calculating the NDVI of each group of data according to the multispectral remote sensing data in step S1, or ; NDVI is the normalized vegetation index, R Green is the red band reflectance, R NIR is the near-infrared band reflectance, R Band8 is the Sentinel-2 8th band reflectance, R Band4 is the Sentinel-2 4th band reflectance; S5, the observation of the spatial and temporal variation law of cyanobacterial bloom: according to the change of the position and intensity of cyanobacterial bloom in different time in each observation area of the Taihu Lake water area, the spatial variation characteristics of cyanobacterial bloom in the Taihu Lake are obtained.

[0009] As an improvement, in step S1, the multispectral remote sensing data is corrected by Sen2Cor plug-in to obtain L2A level data, and then the band data is resampled to 10m by SNAP software.

[0010] As an improvement, the multispectral remote sensing data is greater than 40 groups.

[0011] As an improvement, in step S4, the extraction threshold of NDVI is obtained by the following method: S41, slope analysis is performed on the NDVI value by using the slope tool in Arcgis software, and the natural breakpoint classification method is used to divide the slope into high slope and low slope two categories; S42, reclassifying the slope map, assigning 1 and 0 values to high slope and low slope respectively, and obtaining the NDVI value at the junction of cyanobacterial bloom area and non-cyanobacterial bloom area; S43, statistical analysis is performed on the NDVI value, the NDVI value obtained in step S42 is excluded under the condition of NDVI>0.2 to determine the cyanobacterial bloom pixel, and the non-cyanobacterial bloom pixel interference such as water body is excluded under the condition of NDVI<0; after screening, statistical analysis is performed on the NDVI value of the remaining pixels, the extraction threshold of cyanobacterial pixel is obtained by using the method of mean minus twice standard deviation, and the extraction threshold of NDVI is obtained by averaging the extraction threshold of all remote sensing images.

[0012] As a further improvement, the extraction threshold of NDVI is-0.0693.

[0013] The beneficial effects of the present application are: by analyzing the existing Sentinel-2 remote sensing image and multispectral remote sensing data corresponding to the remote sensing image, accurate actual change data of the Taihu Lake and cyanobacterial bloom in previous years are obtained, and then the reasonable area of the Taihu Lake water area can be conveniently summarized and evaluated to summarize and evaluate the change law of cyanobacterial bloom in the Taihu Lake. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 For the extracted contour of the Taihu Lake water area; Figure 2 For the extraction effect of large-scale cyanobacterial bloom, the specific date of the remote sensing image is marked in the lower right corner of each figure; Figure 3 For the partitioning diagram of the Taihu Lake water area; Figure 4 For the area statistics chart of cyanobacterial bloom in each observation area of the Taihu Lake from 2021 to 2023; Figure 5The average frequency of blue-green algae bloom in each observation area of Taihu Lake in 2021-2023; Figure 6 The variation characteristics of the proportion of blue-green algae bloom area to the total area of Taihu Lake in each observation area in 2021-2023. DETAILED DESCRIPTION

[0015] The application will be further described in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the application and not to limit the scope of the application. In addition, it should be understood that those skilled in the art can make various modifications or changes to the application after reading the content taught by the application, and these equivalent forms also fall within the scope of the claims attached to the present application. EMBODIMENT

[0016] In this study, Sentinel-2 multispectral remote sensing data was downloaded from the European Space Agency website (https: / / dataspace.copernicus.eu / ), with a time range of 2021-2023. L1C level data was processed using the Sen2Cor plug-in for atmospheric correction, etc. to obtain L2A level data, and then the band data was resampled to 10m using SNAP software. In addition, there should be 3-4 remote sensing images per month in the Taihu Lake basin in theory, but in actual research, due to local climate conditions, there are a large number of cloud clusters over the Taihu Lake in some months, making it impossible to analyze the remote sensing images, resulting in fewer available remote sensing images, which may cause incomplete analysis or poor results. However, this phenomenon is unavoidable in the selection of remote sensing images, so these images are still discarded.

[0017] Before extracting blue-green algae blooms in Taihu Lake, in order to prevent non-lake area features from interfering with the extraction results, the lake area to be studied needs to be accurately extracted in advance. A total of 44 images with no cloud cover in the lake area were selected, and NDWI was calculated and the appropriate threshold was selected to completely exclude the land part.

[0018] In this study, the normalized water index (NDWI) was used to extract the water area. NDWI (Normalized Difference Water Index) is a normalized difference value processing of specific bands of remote sensing images. Through this processing method, the water information in the image can be highlighted significantly, so as to more accurately identify and extract the water area. Its calculation formula is: or where, R Green is the green band reflectance, R NIR is the near-infrared band reflectance, R Band3 is the reflectance of the third band of Sentinel-2, and R Band5This represents the reflectance of band 5 of Sentinel-2. Basic information on each band of Sentinel-2 is common knowledge, as detailed in Table 1 of the background technology section. After calculation, a mask for the Taihu Lake area is obtained and applied to subsequent cropping. The water area extraction effect is as follows. Figure 1 As shown.

[0019] NDVI (Normalized Difference Vegetation Index) is a remote sensing index that combines red and near-infrared bands. It is highly sensitive to vegetation information and can be used for monitoring cyanobacterial blooms in lakes. NDVI is calculated based on multispectral data using the following formula: or Wherein, RGreen is the red band reflectance, RNIR is the near-infrared band reflectance, RBand8 is the Sentinel-2 band 8 reflectance, and RBand4 is the Sentinel-2 band 4 reflectance.

[0020] Given the significant differences in spectral characteristics between pixels in cyanobacterial blooms and their boundaries, the cyanobacterial extraction threshold can be determined by statistically analyzing the NDVI slope. Using the slope tool in ArcGIS software, NDVI values ​​were analyzed for slope, and the natural breakpoint classification method was used to categorize slopes into high and low slopes. The slope map was reclassified, assigning values ​​of 1 and 0 to high and low slopes respectively, yielding the NDVI values ​​at the boundary between cyanobacterial bloom and non-cyanobacterial bloom areas. In this data set, pixels identified as cyanobacterial blooms were excluded if NDVI > 0.2, and pixels from non-cyanobacterial bloom areas such as water bodies were excluded if NDVI < 0. After filtering, the NDVI values ​​of the remaining pixels were statistically analyzed. The cyanobacterial pixel extraction threshold for a single image was obtained by subtracting twice the standard deviation from the mean. The extraction thresholds for all images were averaged, resulting in a unified NDVI extraction threshold of -0.0693 based on Sentinel-2 remote sensing imagery. The extraction effect of this method on large-scale cyanobacterial blooms is as follows: Figure 2 As shown.

[0021] Taihu Lake water area can be according to Figure 3 Divided into 7 areas, such as Figure 3 As shown, the diagram illustrates the central lake area and six surrounding edge areas: South Taihu Lake, Northwest Taihu Lake, Xuhu Lake, East Taihu Lake, Meiliang Bay, and Gonghu Bay. Extracting the cyanobacterial bloom area from each of the seven areas allows for analysis of the spatial dynamics of cyanobacterial bloom outbreaks.

[0022] from Figure 4 , Figure 5It can be found that the cyanobacterial bloom area in the south of Taihu Lake often occurs from 2021 to 2023, and there are more serious outbreaks in January, July, and October in 2021, and in August in 2022, with a water bloom area of 150 km 2 Above, and in the long run, from 2021 to 2023, the water bloom area in the south of Taihu Lake showed a downward trend, indicating that the management was effective. Except for January 3, 2021, the water bloom area in the east of Taihu Lake remained small for the rest of the year, almost not exceeding 100 km 2 , and the water bloom area in summer and autumn was slightly higher than that in other periods, indicating that the water environment in the east of Taihu Lake was good, and the cyanobacterial bloom was less frequent, and the water bloom area showed seasonal distribution. The cyanobacterial bloom area in Xuhu was generally small, and there were occasional outbreaks, but it did not exceed 150 km 2 , and generally showed seasonal variation, i.e. higher in summer and autumn than in spring and winter, usually reaching the highest value in July, August or November. Cyanobacterial bloom in the northwest of Taihu Lake occurred frequently, with more serious outbreaks in May 2021, July-November 2022, and July-August 2022, with an area exceeding 200 km 2 , and even exceeding 250 km 2 , but the outbreak of cyanobacterial bloom was significantly reduced in 2023, and the degree was also reduced, indicating that the water quality in the northwest of Taihu Lake was poor, and the pollution was serious, but the situation had improved in recent years, and the water bloom management was effective. The cyanobacterial bloom in Gonghu Bay occurred less frequently, and the size of the single outbreak was small, with an area exceeding 50 km 2 in August and November 2021, and the water bloom area remained low at other times, with good overall water environment quality. The cyanobacterial bloom in Meiliang Bay occurred frequently, although the outbreak size was small, even the highest value in July 2022 was only slightly more than 100 km 2 , but considering that the area of Meiliang Bay itself is small, the water quality and pollution level are still severe. The cyanobacterial bloom in the center of Taihu Lake was generally large, with more serious outbreaks in June 2021, August 2021, November 2021, and June 2022, and the water bloom area even reached 500 km 2The above, the water bloom pollution situation is more serious, but the overall situation in 2023 has improved, and no serious water bloom has occurred. At the same time, by comprehensively analyzing the trend of the area of blue-green algae water bloom in each lake area, the occurrence and diffusion process of blue-green algae water bloom can be obtained. Taking the outbreak of blue-green algae water bloom in July-August 2022 as an example, water bloom first occurred in the northwest lake area and Meiliang Bay, and then spread to the central lake area, Xuhu, Gonghu Bay, and finally accumulated in the south and east Taihu regions, showing a general trend of northwest to southeast. However, due to the influence of weather, there are gaps in the available images, leading to incomplete monitoring of blue-green algae water bloom and difficulty in analyzing its temporal characteristics.

[0023] From Figure 6 it can be found that the area of blue-green algae water bloom in the Xuhu area decreased from 17.3% in 2021 to 12.4% in 2022, but rebounded to 13.2% in 2023, and the area of blue-green algae water bloom in the remaining lake areas accounted for the highest in 2021, and showed a decreasing trend year by year, among which the northwest lake area decreased the most, from 29.1% in 2021 to 25.1% in 2022 and 6.4% in 2023. This indicates that the overall water quality of Taihu has shown a trend of governance from 2021 to 2023.

[0024] According to the “2021 China Ecological Environment Bulletin”, “2022 National Ecological Meteorological Bulletin” and “2023 National Ecological Meteorological Bulletin”, the cumulative area of blue-green algae water bloom in Taihu from 2021 to 2023 was 13185 square kilometers, 11339 square kilometers and 2693 square kilometers, respectively, with a decrease of 14% and 76% respectively. Among them, in 2021, the central lake area and the northwest lake area were lightly polluted, and the eastern coast was moderately eutrophic; in 2022, the frequency of blue-green algae water bloom in the northwest of Taihu was the highest, followed by the north of the central lake area and the southwest coast, and the frequency of blue-green algae water bloom in the east of Taihu was the lowest. The cumulative area of blue-green algae water bloom in autumn was the second largest in the same period in the past five years, and the largest blue-green algae water bloom occurred on October 22; in 2023, it was the smallest area of blue-green algae water bloom in Taihu since 2003, and blue-green algae water bloom occurred in most water areas, but the frequency was relatively small. The water area with high frequency of water bloom is still in the northwest of Taihu. The above data are highly consistent with the analysis results of this study. In recent years, the governance of blue-green algae water bloom in Taihu has achieved remarkable results, and the northwest lake area is still the area where blue-green algae water bloom occurs most frequently, and the remaining areas also need attention.

Claims

1. A method for observing the spatiotemporal variation of cyanobacterial blooms in Taihu Lake based on Sentinel-2 remote sensing images, characterized in that... Includes the following steps: S1. Acquisition of remote sensing data: Acquire Sentinel-2 remote sensing images and the corresponding multispectral remote sensing data; S2. Water Area Contour Extraction: Based on the multispectral remote sensing data from step S1, the NDWI of each data set is calculated, and the land portion is completely excluded according to a threshold to obtain a mask of the Taihu Lake water area, which is then used for subsequent cropping. or NDWI is the normalized water index, R Green R represents the reflectivity in the green band. NIR R represents the near-infrared reflectance. Band3 R represents the reflectivity of the third band of Sentinel-2. Band5 The reflectivity of the 5th band of Sentinel-2; S3. Division of the observation area of ​​Taihu Lake: The outline of Taihu Lake obtained in step S2 is divided into 7 areas, including the central lake area and 6 edge areas surrounding the central lake area. S4. Contour extraction of cyanobacterial blooms: Based on the multispectral remote sensing data in step S1, calculate the NDVI of each group of data. or NDVI is the Normalized Difference Vegetation Index, R Green R represents the reflectivity in the red band. NIR R represents the near-infrared reflectance. Band8 R represents the reflectivity of band 8 of Sentinel-2. Band4 The reflectivity of the 4th band of Sentinel-2; S5. Observation of the spatiotemporal variation of cyanobacterial blooms: Based on the changes in the location and intensity of cyanobacterial blooms in different observation areas of Taihu Lake at different times, the spatial variation characteristics of cyanobacterial blooms in Taihu Lake are obtained.

2. The method for observing the spatiotemporal variation of cyanobacterial blooms in Taihu Lake based on Sentinel-2 remote sensing images according to claim 1, characterized in that, In step S1, the multispectral remote sensing data is corrected using the Sen2Cor plugin to obtain L2A level data, and then the band data is resampled to 10m using SNAP software.

3. The method for observing the spatiotemporal variation of cyanobacterial blooms in Taihu Lake based on Sentinel-2 remote sensing images according to claim 1, characterized in that, The multispectral remote sensing data consists of more than 40 sets.

4. The method for observing the spatiotemporal variation of cyanobacterial blooms in Taihu Lake based on Sentinel-2 remote sensing images according to claim 1, characterized in that, In step S4, the NDVI extraction threshold is obtained by the following method: S41. Use the slope tool in ArcGIS software to perform slope analysis on the NDVI value, and use the natural breakpoint classification method to classify the slope into two categories: high slope and low slope. S42. Reclassify the slope map and assign 1 and 0 values ​​to high slope and low slope respectively to obtain the NDVI value at the boundary between the cyanobacterial bloom area and the non-cyanobacterial bloom area. S43. Perform statistical analysis on the NDVI values. For the NDVI values ​​obtained in step S42, exclude the determined cyanobacterial bloom pixels with NDVI > 0.2 as the condition, and exclude non-cyanobacterial bloom pixels such as water bodies with NDVI < 0 as the condition. After screening, perform statistical analysis on the NDVI values ​​of the remaining pixels, and use the method of subtracting twice the standard deviation from the mean to obtain the cyanobacterial pixel extraction threshold. Then, average the extraction thresholds of all remote sensing images to obtain the NDVI extraction threshold.

5. The method for observing the spatiotemporal variation of cyanobacterial blooms in Taihu Lake based on Sentinel-2 remote sensing images according to claim 4, characterized in that, The extraction threshold for NDVI is -0.0693.

Images