A method for assessing the intensity of urban surface heat island

Abstract

This invention provides a method for assessing the intensity of urban surface heat islands. It acquires multi-band satellite imagery of the target area during the study period, delineates water bodies and urban built-up areas, selects a portion of the water bodies as reference base stations, extracts seasonal and trend factors based on long-term observed surface temperature data of the reference base stations, corrects the surface temperature data of the reference base stations to obtain a corrected temperature, and compares this with the surface temperature of the urban built-up areas to obtain the intensity of the urban surface heat island. This quantifies the seasonal fluctuation trend of the temperature in the reference base station areas, making the spatial distribution of urban thermal environment and thermal risk clearer, more intuitive, and easier to locate, thus helping to reduce assessment errors caused by seasonal and background climate changes.

Description

Technical Field

[0001] This invention relates to the field of environmental research, and more specifically, to a method for assessing the intensity of urban surface heat islands. Background Technology

[0002] The urban heat island effect is one of the most obvious climate changes brought about by urbanization. Besides exacerbating public health risks, the urban heat island effect also indirectly leads to adverse effects such as increased energy consumption and aggravated air pollution, posing a significant challenge to urban sustainability. With further urban expansion / densification and the coupling effect of global warming, these adverse effects will escalate, generating more complex and difficult-to-manage compound and cascading risks. Urban surface heat island intensity, due to its simple calculation and comparability, is one of the most widely used key indicators in current multi-scale heat island research. Accurate quantification of urban surface heat island intensity is crucial for understanding and addressing the processes of thermal environmental change.

[0003] The quantification of urban surface heat island intensity depends on the definition of the reference area. Existing methods for defining reference areas are inconsistent and affected by factors such as the complexity of internal land cover and seasonal fluctuations. These factors reduce the universality and accuracy of urban heat island intensity quantification, leading to cognitive biases and even misconceptions in the assessment of the urban heat island effect, and creating difficulties for subsequent policy formulation and adaptation mitigation measures. Summary of the Invention

[0004] This invention provides a method for assessing the intensity of urban surface heat islands, which improves upon the lack of universality and accuracy in the existing methods for quantifying urban heat island intensity, thereby enhancing the universality and accuracy of urban heat island intensity quantification.

[0005] To achieve the above objectives, the present invention provides a method for assessing the intensity of urban surface heat islands, comprising the following steps:

[0006] S1. Acquire satellite multi-band imagery of the target area during the research period to obtain surface temperature data;

[0007] S2. Divide the satellite multi-band data imagery into at least water bodies and urban built-up areas, wherein the surface temperature of the urban built-up areas is the assessment temperature;

[0008] S3. Select a portion of the water body as the reference benchmark station area; the temperature of the water body is relatively stable compared to the land, which is beneficial to improving the accuracy of the assessment method.

[0009] S4. Based on the surface temperature data of the reference station area during the research period, extract seasonal and trend factors, and correct the surface temperature data of the reference station area based on the seasonal and trend factors to obtain a reference temperature. Extracting the seasonal and trend factors and correcting the surface temperature accordingly can quantify the seasonal fluctuation trend of the temperature in the reference station area and reduce the additional impact of macro-climate change, which helps to reduce the assessment error caused by seasonal or background climate change.

[0010] S5. The difference between the assessed temperature and the reference temperature is the intensity of the surface heat island.

[0011] As an optional technical solution, step S2 includes:

[0012] S201. Select training samples based on the satellite multi-band data imagery;

[0013] S202. Construct a random forest model using the training samples;

[0014] S203. Apply the random forest model to perform supervised land cover classification on the satellite multi-band data image to obtain land cover data;

[0015] S204. Based on land cover data, the satellite multi-band data imagery is divided into construction land, water bodies, agricultural and forestry land, and other land uses;

[0016] S205. Extract the construction land pixels from the land cover data based on the urban clustering algorithm;

[0017] S206. Use the extracted built-up land pixels to identify the urban built-up area. By identifying the urban built-up area, the sampling location for the assessed temperature can be clearly identified. On the other hand, the influence range of the urban heat island can be delineated based on the scope of the urban built-up area, and the water bodies within the influence range can be discharged from the reference station area, thereby improving the reliability of the reference temperature.

[0018] As an optional technical solution, step S3 includes:

[0019] S301. Select the water body as the reference benchmark station area;

[0020] S302. Verify the stability of the reference station area by combining different remote sensing data sources;

[0021] S303. The temperature fluctuations and differences from other cover areas of the reference benchmark station area are examined using analysis of variance and statistical characteristics to test its robustness. The reference benchmark station area is then adjusted based on the test results. Through cross-validation of different data sources and the selection of statistical methods, a subset of the most stable water bodies can be selected as the reference benchmark station area from all candidate water bodies, which helps improve the accuracy of urban heat island assessment.

[0022] As an optional technical solution, the screening conditions in step S301 include:

[0023] Condition 1: The water body is relatively still and has an area greater than or equal to 0.01 square kilometers to ensure its high heat capacity and thermal stability due to the scale effect;

[0024] Condition 2: The water body is located within the administrative division of the target area and within the urban built-up area heat island footprint range to comply with urban planning and reduce the impact of urbanization on the water body;

[0025] Condition 3: The elevation of the water body is within ±50 meters of the elevation of the urban built-up area, in order to reduce the impact of elevation.

[0026] Condition 4: The extent of the water body shall not change by more than 10% within five years to ensure its spatial stability and planning availability;

[0027] The heat island footprint refers to the continuous spatial range from the urban built-up area to its fifth equal-area buffer zone.

[0028] As an optional technical solution, step S302 includes: when the set of reference base stations is lower than the hard reference value, detecting, removing and optimizing outliers and irregularities;

[0029] The reference base station set being lower than the hard reference value includes at least the following: an ANOVA p-value less than 0.05 and a standard deviation less than other coverages; a Spearman rank correlation coefficient greater than 0.7 and / or a significance p-value less than 0.05.

[0030] As an optional technical solution, step S4 includes:

[0031] S401. Noise processing is applied to the surface temperature within the reference station area;

[0032] S402. Based on historical remote sensing data, record long-term temperature data covering different seasons in the reference station area to form a temperature time series;

[0033] S403. Extract the seasonal and trend components from the temperature time series using time series decomposition; wherein the trend component can reflect the long-term linear trend of climate change in the target area and / or potential acceleration or deceleration effects, which helps to express the relatively complex climate change process in the target area.

[0034] S404. Based on the extracted seasonal components and trend components, construct a periodic temperature model for the benchmark water body, the model being used to describe the annual temperature variation;

[0035] S405. Based on the current seasonal temperature value predicted by the model, calculate the temperature correction factor that should be applied in each time period;

[0036] S406. The reference temperature is obtained by adding the observed temperature of the water body during the study period to the temperature correction factor.

[0037] As an optional technical solution, the model is as follows:

[0038]

[0039] Where a*t 2 +b*t represents the trend component. Let i be the seasonal component, and let LST be the random component. t_base Let be the reference temperature for time period t, a be the quadratic coefficient, b be the linear coefficient, c be the seasonal amplitude coefficient, P be the period, d be the phase shift, and t be a subset of the study period. The above expression of the trend component introduces the quadratic coefficient 'a' and the linear coefficient 'b' to reflect the potential acceleration or deceleration effects of climate change in the target region and the long-term linear trend of climate change, respectively. This makes the model more flexible and capable of expressing relatively complex change processes.

[0040] As an optional technical solution, the temperature correction factor is calculated as follows:

[0041] C LST =LST t_base -LST t_real

[0042] C LST LST is the correction factor for time period t. t_base LST is the reference temperature at time t. t_real The observed temperature of the water body. The reference station area typically includes multiple water bodies, and the correction factor can intuitively reflect the amount of correction applied to the temperature of a single water body, facilitating researchers' understanding of the operation of the assessment method and the development of corresponding strategies.

[0043] As an optional technical solution, step S401 includes:

[0044] S4011. Combining the local mean and standard deviation of the reference station area, noise identification and screening are performed based on the dynamic noise threshold.

[0045] S4012. Apply a Gaussian filtering smoothing algorithm to smooth the local area data of the reference station area to reduce the impact of random noise.

[0046] S4013. Calculate the temperature difference between each temperature point and its neighboring data. Points with abnormally large temperature differences are marked as noise candidates.

[0047] As an optional technical solution, step S5 is as follows:

[0048]

[0049] Where SUHII represents the heat island intensity, and LST represents the heat island intensity. Object The surface temperature of the urban built-up area is denoted as .

[0050] One or more technical solutions provided by this invention have at least the following technical effects or advantages:

[0051] The urban surface heat island intensity assessment method of the present invention acquires satellite multi-band imagery of the target area during the study period, delineates water bodies and urban built-up areas based on this data, selects a portion of the water bodies as reference base station areas, extracts seasonal and trend factors based on long-term observed surface temperature data of the reference base station areas, corrects the surface temperature data of the reference base station areas to obtain corrected temperatures, and compares these temperatures with the surface temperatures of the urban built-up areas to obtain the surface heat island intensity of the urban built-up areas. This quantifies the seasonal fluctuation trend of the temperature of the reference base station areas, making the spatial distribution of urban thermal environment and thermal risk clearer, more intuitive, and easier to locate, thus helping to reduce assessment errors caused by seasonal changes. Attached Figure Description

[0052] The accompanying drawings, which are provided to further illustrate embodiments of the invention and constitute a part of this invention, are not intended to limit the scope of the invention.

[0053] Figure 1 This is a flowchart of the urban surface heat island intensity assessment method in this invention;

[0054] Figure 2 This is an example image of the surface temperature of the case area obtained by inversion based on satellite multi-band data imagery in an embodiment of the present invention;

[0055] Figure 3 This is a schematic diagram of the land cover status of the target area in an embodiment of the present invention;

[0056] Figure 4 This is a reference base station area identification image in an embodiment of the present invention;

[0057] Figure 5 This is a schematic diagram illustrating the surface temperature changes of the reference station area and other land cover in an embodiment of the present invention.

[0058] Figure 6 This is a graph showing the reference temperature in an embodiment of the present invention;

[0059] Figure 7 This is an example image of the target region based on the surface heat island intensity results from the reference station area. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] Example 1

[0062] This embodiment provides a method for assessing the intensity of urban surface heat islands, including the following steps:

[0063] S1. Acquire satellite multi-band imagery of the target area during the research period to obtain surface temperature data;

[0064] S2. Divide the satellite multi-band data imagery into at least water bodies and urban built-up areas, wherein the surface temperature of the urban built-up areas is the assessment temperature;

[0065] S3. Select a portion of the water body as the reference benchmark station area; the temperature of the water body is relatively stable compared to the land, which is beneficial to improving the accuracy of the assessment method.

[0066] S4. Based on the surface temperature data of the reference station area during the research period, extract seasonal factors, and correct the surface temperature data of the reference station area based on the seasonal factors to obtain the reference temperature;

[0067] S5. The difference between the assessed temperature and the reference temperature is the intensity of the surface heat island.

[0068] Step S4 introduces seasonal and trend factors to correct the surface temperature data of the reference station area, establishing a cross-seasonal correction mechanism. This allows the assessment method of the present invention to dynamically adapt to the natural periodic changes in water temperature, thereby more accurately assessing the intensity of the urban heat island. By extracting periodic and trend components, the influence of seasons on the fluctuation trend of reference temperature and the background climate can be quantified. This helps reduce assessment errors caused by seasonal variations and macroeconomic trends.

[0069] Water temperature varies across different periods due to a variety of factors. Periodic correction mechanisms, by adjusting model parameters in real time, ensure the accuracy and consistency of assessments across different periods. This helps to obtain comparable urban heat island intensity assessment results over different time periods.

[0070] Periodic correction mechanisms can provide accurate information on seasonal variations in water body temperature, which helps optimize resource allocation. For example, during the hot summer months, it may be necessary to increase monitoring and mitigation measures for urban heat islands; while during the cold winter months, these measures can be reduced accordingly. Through periodic corrections, relevant strategies can be formulated and implemented more scientifically.

[0071] Optionally, step S1 includes:

[0072] S101. Acquire satellite multi-band imagery of the target area during the research period, including digital signal values;

[0073] S102. Preprocess satellite multi-band data images, including mosaic correction, cropping, cloud and shadow masking, and image registration.

[0074] S103. Convert the digital signal value into radiance;

[0075] S104. Correct the radiance data by combining atmospheric parameters, thereby eliminating errors caused by the scattering, absorption and reflection of radiation on the Earth's surface by the atmosphere;

[0076] S105. Apply the radiative transfer equation to calculate the surface temperature, and use the Planck function to calculate the actual temperature value.

[0077] Example 2

[0078] Based on Embodiment 1, step S2 includes:

[0079] S201. Select training samples based on the satellite multi-band data imagery;

[0080] S202. Construct a random forest model using the training samples;

[0081] S203. Apply the random forest model to perform supervised land cover classification on the satellite multi-band data image to obtain land cover data;

[0082] S204. Based on land cover data, the satellite multi-band data imagery is divided into construction land, water bodies, agricultural and forestry land, and other land uses;

[0083] S205. Extract the construction land pixels from the land cover data;

[0084] S206. Use the extracted built-up land pixels to identify the urban built-up area. By identifying the urban built-up area, the sampling location for the assessed temperature can be clearly identified. On the other hand, the influence range of the urban heat island can be delineated based on the scope of the urban built-up area, and the water bodies within the influence range can be discharged from the reference station area, thereby improving the reliability of the reference temperature.

[0085] The criteria for classifying land cover types can be found in the table below:

[0086]

[0087] As an optional technical solution, step S3 includes:

[0088] S301. Select the water body as the reference benchmark station area;

[0089] S302. Verify the stability of the reference station area by combining different remote sensing data sources;

[0090] S303. The temperature fluctuations and differences from other cover areas of the reference benchmark station area are examined using analysis of variance and statistical characteristics to test its robustness. The reference benchmark station area is then adjusted based on the test results. Through cross-validation of different data sources and the selection of statistical methods, a subset of the most stable water bodies can be selected as the reference benchmark station area from all candidate water bodies, which helps improve the accuracy of urban heat island assessment.

[0091] The aforementioned remote sensing data sources can include MODIS surface temperature data, ECOSTRESS surface temperature data, UAV thermal infrared imagery, etc.

[0092] As an optional technical solution, the screening conditions in step S301 include:

[0093] Condition 1: The water body is relatively still and has an area greater than or equal to 0.01 square kilometers to ensure its high heat capacity and thermal stability due to the scale effect;

[0094] Condition 2: The water body is located within the administrative division of the target area and within the urban built-up area heat island footprint range to comply with urban planning and reduce the impact of urbanization on the water body;

[0095] Condition 3: The elevation of the water body is within ±50 meters of the elevation of the urban built-up area, in order to reduce the impact of elevation.

[0096] Condition 4: The extent of the water body shall not change by more than 10% within five years to ensure its spatial stability and planning availability;

[0097] The heat island footprint refers to the continuous spatial range from the urban built-up area to its fifth equal-area buffer zone.

[0098] As an optional technical solution, step S302 includes: when the set of reference base stations is lower than the hard reference value, detecting, removing and optimizing outliers and irregularities;

[0099] The reference benchmark area set being lower than the hard reference value includes at least the following: a p-value of less than 0.05 and a standard deviation lower than other land cover; a Spearman rank correlation coefficient greater than 0.7 and a significance p-value less than 0.05.

[0100] As an optional technical solution, step S4 includes:

[0101] S401. Noise processing is applied to the surface temperature within the reference station area;

[0102] S402. Based on historical remote sensing data, record long-term temperature data covering different seasons in the reference station area to form a temperature time series;

[0103] S403. Extract the seasonal and trend components from the temperature time series using time series decomposition; wherein the trend component can reflect the long-term linear trend of climate change in the target area and / or potential acceleration or deceleration effects, which helps to express the relatively complex climate change process in the target area.

[0104] S404. Based on the extracted seasonal components and trend components, construct a periodic temperature model for the benchmark water body, the model being used to describe the annual temperature variation;

[0105] S405. Based on the current seasonal temperature value predicted by the model, calculate the temperature correction factor that should be applied in each time period;

[0106] S406. The reference temperature is obtained by adding the observed temperature of the water body during the study period to the temperature correction factor.

[0107] Optionally, step S401 includes:

[0108] S4011. Noise identification and screening are performed based on dynamic noise threshold by combining local mean and standard deviation;

[0109] S4012. Apply a Gaussian filtering smoothing algorithm to smooth the data in the base region and reduce the impact of random noise; the expression for the Gaussian filtering smoothing algorithm is:

[0110]

[0111] Where K(x, y) is the weight of the Gaussian kernel (x, y) at the position; k is half the size of the kernel, and in this embodiment, a 3*3 kernel is used, i.e., k = 1; I(i+x, j+y) is the pixel value at the corresponding position in the original image.

[0112] S4013. Calculate the temperature difference between each temperature point and its neighboring data. Points with abnormally large temperature differences are marked as noise candidates to improve detection accuracy.

[0113] As an optional technical solution, the model is as follows:

[0114]

[0115] Where a*t 2 +b*t represents the trend component. Let i be the seasonal component, and let LST be the random component. t_base Let be the reference temperature for time period t, a be the quadratic coefficient, b be the linear coefficient, c be the seasonal amplitude coefficient, P be the period, d be the phase shift, and t be a subset of the study period. The above expression of the trend component introduces the quadratic coefficient 'a' and the linear coefficient 'b' to reflect the potential acceleration or deceleration effects of climate change in the target region and the long-term linear trend of climate change, respectively. This makes the model more flexible and capable of expressing relatively complex change processes.

[0116] As an optional technical solution, the temperature correction factor is calculated as follows:

[0117] C LST =LST t_base -LST t_real

[0118] C LST LST is the correction factor for time period t. t_base LST is the reference temperature at time t. t_realThe observed temperature of the water body. The reference station area typically includes multiple water bodies, and the correction factor can intuitively reflect the amount of correction applied to the temperature of a single water body, facilitating researchers' understanding of the operation of the assessment method and the development of corresponding strategies.

[0119] As an optional technical solution, after step S401 and before step S402, the following steps are also included:

[0120] S4011. Combining the local mean and standard deviation of the reference station area, noise identification and screening are performed based on the dynamic noise threshold.

[0121] S4012. Apply a Gaussian filtering smoothing algorithm to smooth the local area data of the reference station area to reduce the impact of random noise.

[0122] S4013. Calculate the temperature difference between each temperature point and its neighboring data. Points with abnormally large temperature differences are marked as noise candidates.

[0123] As an optional technical solution, step S5 is as follows:

[0124]

[0125] Where SUHII represents the heat island intensity, and LST represents the heat island intensity. Object The surface temperature of the urban built-up area is denoted as .

[0126] Furthermore, in practice, urban thermal environment classification standards can be negotiated and stipulated independently. The thermal environment conditions of each assessment unit are compared according to the standards, and optimized designs are carried out accordingly. The following suggestions are provided: Assessment units facing severe heat island effects should be marked in red to increase the priority of remediation. Planners and designers should employ one or more cooling strategies for combined cooling based on local climate simulations. After the measures are implemented, surface heat island intensity assessments should be continuously and regularly repeated to monitor changes in the thermal environment and the effectiveness of the remediation measures.

[0127] Example 3

[0128] This embodiment takes a city as an example and conducts an urban heat island intensity assessment using the method described in Embodiment 2. First, high-quality images from the Landsat remote sensing image dataset (Landsat-8 and Landsat-9) since 2017, with cloud cover controlled to be below 20%, are selected as the raw data, totaling 81 images. Preprocessing steps, including mosaic correction, cropping, cloud and shadow masking, and image registration, are performed on the raw dataset. Finally, the radiative transfer equation is applied to calculate the land surface temperature. Figure 2 The study showcases surface temperature data for a case study area obtained from satellite multi-band imagery inversion over a period of time.

[0129] like Figure 3 As shown, based on the construction land area in the 2023 land cover data, the urban built-up area is identified by the urban clustering algorithm, where the scanning grid is 100 meters and the clustering distance is 200 meters.

[0130] Reference benchmark station areas were selected based on multi-source data, including land cover data, administrative division data, and elevation data. The results are as follows: Figure 4 As shown, a total of 19 water bodies meeting the criteria were selected within the case study area to form a set of reference station areas.

[0131] like Figure 5 As shown in the table, partial time data were selected as validation samples to statistically analyze the surface temperature conditions of different land cover types and the reference station area within the target region. Analysis of variance based on the mean (Avg) revealed significant differences among different land cover types (p<0.05). Specifically, the standard deviation (Std) of temperature within the reference station area was significantly smaller than that of other land cover types, indicating greater internal homogeneity and reduced influence from outliers. Time series data also showed that the surface temperature changes in the reference station area under different climatic backgrounds were smaller and more robust compared to other land cover types.

[0132]

[0133] The reliability of the reference station area compared to other land cover was further verified by combining downscaled MODIS land surface temperature data, as shown in the table below. The results show that there are significant differences in land surface temperature between the reference station area and other land cover types, with a smaller standard deviation. This verifies the reliability and stability of the reference station area. Furthermore, there is a high correlation between the MODIS land surface temperature time series of the reference station area and Landsat data, with a Spearman correlation coefficient of 0.88 and a significance p-value less than 0.05. This indicates that the reference station area has strong applicability to different land surface temperature data.

[0134]

[0135] like Figure 6 As shown, based on 81 surface temperature images covering different seasons from 2017 to 2024, a temperature time series was formed. After smoothing the anomalies in the water bodies within the reference station area, a sine curve mathematical model describing the annual temperature change was formed, namely the periodic temperature model of the reference water body mentioned above, which quantifies the fluctuation trend of the seasonal influence on the reference temperature.

[0136] A correction factor is calculated based on the difference between the model and the actual temperature, which is used to calculate the intensity of the urban heat island in a region. Examples of calculation results for some dates are shown below. Figure 7 As shown.

[0137] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0138] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for assessing the intensity of urban surface heat islands, characterized in that, Includes the following steps: S1. Acquire satellite multi-band imagery of the target area during the research period to obtain surface temperature data; S2. Divide the satellite multi-band data imagery into at least water bodies and urban built-up areas, wherein the surface temperature of the urban built-up areas is the assessment temperature; S3. Select a portion of the aforementioned water bodies as a reference benchmark station area. The selection criteria include: Condition 1: The water body is relatively still, and its area is greater than or equal to 0.01 square kilometers; Condition 2: The water body is located within the administrative division of the target area and outside the urban built-up area heat island footprint range; Condition 3: The elevation of the water body is within ±50 meters of the elevation of the urban built-up area; Condition 4: The range of the water body shall not change by more than 10% within five years; wherein, the heat island footprint is the continuous spatial range from the urban built-up area to its fifth equal-area buffer zone; S4. Based on the surface temperature data of the reference station area during the research period, extract seasonal and trend factors, and correct the surface temperature data of the reference station area based on the trend factors and the seasonal factors to obtain the reference temperature; S5. The difference between the assessed temperature and the reference temperature is the surface heat island intensity; Step S4 includes: S401. Noise processing is applied to the surface temperature within the reference station area; S402. Based on historical remote sensing data, record long-term temperature data covering different seasons in the reference station area to form a temperature time series; S403. Extract the seasonal and trend components from the temperature time series using time series decomposition; S404. Based on the extracted seasonal components and trend components, a periodic temperature model of the reference water body is constructed, wherein the periodic temperature model is used to describe the temperature variation throughout the year. S405. Based on the current seasonal temperature value predicted by the periodic temperature model, calculate the temperature correction factor that should be applied in each period. S406. The reference temperature is obtained by adding the observed temperature of the water body during the study period to the temperature correction factor; The periodic temperature model is as follows: Where a*t 2 +b*t represents the trend component. Let i be the seasonal component, and let LST be the random component. t_base Let t be the reference temperature for time period t, a be the quadratic coefficient, b be the linear coefficient, c be the seasonal amplitude coefficient, P be the period, d be the phase shift, and t be a subset of the study period. The temperature correction factor is calculated as follows: C LST =LST t_base -LST t_real C LST LST is the correction factor for time period t. t_base LST is the reference temperature at time t. t_real The observed temperature of the water body.

2. The method for assessing the intensity of urban surface heat islands according to claim 1, characterized in that, Step S2 includes: S201. Select training samples based on the satellite multi-band data imagery; S202. Construct a random forest model using the training samples; S203. Apply the random forest model to perform supervised land cover classification on the satellite multi-band data image to obtain land cover data; S204. Based on land cover data, the satellite multi-band data imagery is divided into construction land, water bodies, agricultural and forestry land, and other land uses; S205. Extract the construction land pixels from the land cover data; S206. Using the extracted construction land pixels, an urban clustering algorithm is used to identify the urban built-up area.

3. The method for assessing the intensity of urban surface heat islands according to claim 1, characterized in that, Step S3 includes: S301. Select the water body as the reference benchmark station area; S302. Verify the stability of the reference station area by combining different remote sensing data sources; S303. Combine analysis of variance and statistical characteristics to test the temperature fluctuations of the reference station area and the differences from other cover, so as to test its robustness, and adjust the reference station area according to the test results.

4. The method for assessing the intensity of urban surface heat islands according to claim 3, characterized in that, Step S302 includes: when the set of reference base stations is lower than the hard reference value, detecting, removing and optimizing outliers and abnormal values; The reference base station set being lower than the hard reference value includes at least the following: an ANOVA p-value less than 0.05 and a standard deviation less than other coverages; a Spearman rank correlation coefficient greater than 0.7 and a significance p-value less than 0.

05.

5. The method for assessing the intensity of urban surface heat islands according to claim 1, characterized in that, Step S401 includes: S4011. Combining the local mean and standard deviation of the reference station area, noise identification and screening are performed based on the dynamic noise threshold. S4012. Apply a Gaussian filtering smoothing algorithm to smooth the local area data of the reference station area to reduce the impact of random noise. S4013. Calculate the temperature difference between each temperature point and its neighboring data. Points with abnormally large temperature differences are marked as noise candidates.

6. The method for assessing the intensity of urban surface heat islands according to claim 1, characterized in that, Step S5 is as follows: Where SUHII represents the heat island intensity, and LST represents the heat island intensity. Object The surface temperature of the urban built-up area is denoted as .

Images