CN109903234B - Quantitative description and multi-scale feature analysis method for urban thermal landscape - Google Patents

Abstract

The present invention claims to protect a method for quantitative description and multi-scale feature analysis of urban thermal landscape, including: preprocessing Landsat‑8 OLI/TIRS remote sensing images through existing methods such as radiation calibration, atmospheric correction, geometric correction, etc., after preprocessing The processed images are inverted into land surface temperature images using existing methods. For the study area, the improved "mean-standard deviation" method was used to divide the surface temperature images into thermal landscapes, and the landscape index was calculated at the three levels of patch, type and landscape, and the spatial pattern of urban thermal landscape was analyzed based on the index results. Change the scale of two or more research areas, including spatial granularity transformation and spatial amplitude change, calculate the selected landscape index on the spatial granularity and amplitude and analyze its multi-scale characteristics, and finally for the multi-scale features in different research areas Comparative analysis of changing trends. This method can be used to analyze the urban thermal landscape of a single or multiple study areas, and the analysis of this method is more comprehensive and specific.

Description

A quantitative description and multi-scale feature analysis method of urban thermal landscape

technical field

The invention belongs to the field of urban heat island and spatial structure analysis. Specifically, it involves a landscape index-based quantitative description method for analyzing the spatial distribution characteristics and multi-scale characteristics of urban heat islands in different research areas.

Background technique

The Urban Heat Island Effect (UHI) refers to a climate phenomenon in which the temperature in urban areas is higher than that in suburban areas due to factors such as human activities, high heat storage bodies such as buildings and roads, and the reduction of green spaces and water bodies. The urban heat island effect aggravates the frequency of high temperature in cities and the possibility of high temperature disasters, which makes the environmental quality of urban areas decline and seriously affects the life of urban residents and the sustainable development of cities. It is generally believed that the ecological process determines the landscape pattern, and the landscape pattern will in turn affect the ecological process. Therefore, analyzing the interaction between a city's landscape pattern and the urban heat island process is important for understanding the urban heat island effect and mitigating the urban heat island. meaning.

Since the 1950s, the urban heat island effect has gradually become one of the hot issues in the study of urban climatology. Scholars at home and abroad have used meteorological observation, ground remote sensing monitoring, numerical simulation and other research methods to study the formation mechanism of urban heat island, the intensity of urban heat island effect and Extensive research has been carried out on the characteristics of temporal and spatial changes, the hazards of the urban heat island effect, and the mitigation measures of the urban heat island effect. Landscape ecology is a relatively young and widely applied branch of ecology, mainly focusing on the type, number, spatial distribution and pattern of landscape components. The coupling of landscape pattern and ecological process is the core research field of landscape ecology. Landscape pattern not only reflects the result of the interaction of various ecological processes at different time and space scales, but also determines the distribution of various natural environmental factors in landscape space. And combinations, thereby restricting various ecological processes. As a representative ecological environment effect of urbanization, the heat island effect is one of the most direct consequences of the evolution of urban landscape patterns. At present, the research on urban heat island based on landscape pattern is mainly carried out from three aspects: distribution characteristics, particle size effect, and spatial-temporal evolution. Among them, the clear division of land use and land cover (LULC) on the surface is the key to landscape and urban development. The most common practice in heat island analysis and indeed in other studies in landscape ecology. Scholars such as Chen L, Xu S, Weng Q, etc. used the traditional landscape index to quantitatively describe the spatial distribution characteristics of thermal landscape; Scholars Shen Weijun, Xu Jianhua, Meng Chen, etc. Relevant studies have been carried out and many conclusions have been drawn; Yue Wenze, Bai Y, Deng Rui and other scholars have analyzed the temporal and spatial evolution of urban heat island landscape patterns from the perspective of time, and have drawn many conclusions that are helpful for urban planning and mitigation of urban heat island. effect conclusions. In addition, some scholars have also conducted relevant discussions on the spatial amplitude effect of urban thermal landscape patterns, in order to analyze the internal formation mechanism and spatial distribution characteristics of urban heat island landscape patterns from different perspectives and more comprehensively.

Although significant progress has been made in research on the urban heat island effect, the current understanding of the urban heat island effect is still incomplete. Most scholars' research on landscape pattern and urban heat island has finally been transformed into an analysis of the relationship between LULC and LST or LULC and UHI landscape, lacking a direct analysis of landscape pattern of UHI landscape divided by LST. The research on the scale effect is only for a single city or a single region, lacking the scale effect analysis based on two or more research areas. Moreover, the research area selected by most scholars is a rectangle, and it is generally believed that the information carried by the circular research area is more extensive and continuous. Whether there is a shifting feature of the scale effect in the two research areas with completely different topographical structures, that is, multi-scale features in the Whether there are similar trends in different study areas. The present invention aims at proposing a method for quantitatively analyzing spatial distribution characteristics and multi-scale characteristics of urban thermal landscape based on landscape index.

Contents of the invention

The present invention aims to solve the above problems of the prior art. A more comprehensive and specific method for analyzing multi-scale characteristics of urban urban heat island landscape based on landscape index is proposed. Technical scheme of the present invention is as follows:

A method for quantitative description and multi-scale feature analysis of urban thermal landscape, comprising the following steps:

Step S1, obtaining Landsat-8OLI/TIRS remote sensing satellite image data in the study area;

Step S2, remote sensing image data preprocessing: perform preprocessing on the Landsat-8OLI/TIRS remote sensing image data obtained in step S1, including radiometric calibration, atmospheric correction, geometric correction and cropping;

Step S3, inverting the surface temperature: using the classical radiative transfer equation method to invert the surface temperature of the remote sensing image data preprocessed in step S2, to obtain the surface temperature data of the study area;

Step S4, Classification of Thermal Landscape: Use random sampling to obtain the highest surface temperature value in the vegetation area, and then use the mean-standard deviation method to formulate thermal landscape classification standards;

Step S5, selection of landscape pattern index: According to the geographical environment and research needs of the study area, select the required index from the three levels of patch, category, and landscape, and substitute the thermal landscape data obtained in step S4 into the calculation and get the result , and quantitatively analyze the spatial distribution characteristics of the urban heat island according to the results;

Step S6, scale change: perform granularity transformation on the thermal landscape data obtained in step S4 from the spatial granularity, and change the amplitude from the spatial amplitude;

Step S7, multi-scale feature analysis, according to the landscape pattern index selected in step S5 and the scale change in step S6, calculate the spatial granularity and spatial extent of the data at the three levels of patch, category, and landscape from the perspective of spatial granularity and spatial extent. According to the index results, the multi-scale characteristics of urban heat island landscape are obtained according to the response relationship of urban heat island landscape analysis index represented by landscape index with the change of spatial scale.

Further, the step S2 remote sensing image data preprocessing specifically includes: using existing software including ENVI 5.3, using other high-resolution satellite images as reference images, selecting 30 to 40 road intersections, river bends and intersections The feature point with the same name is used as the control point, and all bands in the Landsat 8OLI/TIRS image including the thermal infrared band are subjected to a quadratic polynomial geometric fine correction, and then the radiometric calibration tool and FLAASH tool in the ENVI 5.3 software are used to correct each band Carry out radiometric calibration and atmospheric correction, obtain preprocessed data and cut the data according to requirements to obtain the final data of the study area.

Further, the step S3 uses the classical radiation transfer equation method to invert the surface temperature, including the steps: according to the data obtained in the step S2, first divide the object categories in the research area into three major categories: natural surface, town, and water body According to the normalized difference vegetation index (Normalized Difference Vegetation Index, NDVI) inversion of the surface emissivity, the water body pixel is often a single water body, so the emissivity ε between the natural surface and the town is mainly considered, and the satellite pixel scale ε can be calculated as the formula: ε＝P _V R _V ε _V +(1-P _V )R _X ε _X +dε In the formula, P _V represents the vegetation coverage, R _V is the temperature ratio of vegetation, and ε _V is the temperature ratio of vegetation. Specific emissivity, R _X is the temperature ratio of bare soil or building surface, ε _X is the specific emissivity of bare soil or building surface, dε is the thermal radiation interaction, according to the vegetation coverage, estimate the temperature ratio of ground objects, vegetation, bare Soil and building surface temperature ratio; secondly, estimate the influence of the atmosphere on the surface thermal radiation, and calculate the thermal infrared radiation brightness value received by the satellite sensor: L _λ =[εB(T _surface )+(1-ε)L _down ]τ+ L _upper , where, L _λ represents the thermal infrared radiance received by the satellite sensor; ε represents the specific emissivity of the surface; τ represents the transmittance of the atmosphere in the thermal infrared band; T _surface represents the true surface temperature (K); B (T _surface ) represents the radiance of a black body with temperature T _s ; Lu _upper and L _down represent the upward and downward radiance of the atmosphere respectively, and finally use Planck's formula 1.3 to obtain the surface temperature T _sruface : T _surface ＝K ₂ /[ln (1+K ₁ /B(T _surface ))], where K ₁ and K ₂ are constants, which can be obtained from the header file of the corresponding data.

Further, the step S4 uses random sampling to obtain the highest surface temperature value in the vegetation area, and then uses the mean-standard deviation method to formulate thermal landscape classification standards, specifically including:

其中，R_LST为计算出的温度阈值范围，

为研究区所选植被样本区域的最高温度，SD为地表温度标准差，n为标准差的倍数，依据分类公式将研究区域划分为4种热力景观类型，依次为无热岛Non heat island、弱热岛Weak heat island、热岛Heat island和强热岛Strong heatisland，阈值表示如下公式所述：For the classification of single-band grayscale images, the segmentation interval is selected as the maximum value and minimum value of the surface temperature, and the surface temperature image is divided into grades to reflect the type of thermal landscape. The classification formula is as follows:

where R _LST is the calculated temperature threshold range,

is the maximum temperature of the selected vegetation sample area in the research area, SD is the standard deviation of the surface temperature, and n is the multiple of the standard deviation. According to the classification formula, the research area is divided into four thermal landscape types, which are Non heat island and weak heat island Weak heat island, heat island Heat island and strong heat island Strong heat island, the threshold is expressed as described in the following formula:

Further, the selection of the landscape pattern index in the step S5 specifically includes: using Fragstats 4.2 to calculate the landscape pattern index, combined with the landscape characteristics of the research area, selecting the required index from the three levels of patch, category, and landscape for calculation. The index calculation formula For its meaning, see Fragstats 4.2 software instructions and related literature. According to the results of the index, the spatial distribution characteristics of the urban thermal landscape are analyzed from the aspects of area, density, shape, convergence and diversity.

Further, from the five aspects of area, density, shape, convergence and diversity, the maximum patch index LPI, the percentage of patch area occupied by the landscape PLAND, the number of patches NP, the patch density PD, the landscape shape index LSI, Aggregation Index AI, Contagion Index CONTAG, Shannon Diversity Index SHDI.

Further, according to the thermal landscape data obtained in step S4, the step S6 performs granular transformation on the spatial granularity, and performs amplitude changes on the spatial amplitude, specifically including:

The total area of the study area determines the spatial extent of the study. Using the multi-scale analysis method of landscape pattern that changes the spatial scale, the original resolution of the study area is 30m as the starting point, and the image is resampled to 300m according to the step size of 30m, and then according to Resample to 600m with a step size of 60m, and finally resample to 960m with a step size of 120m to complete the transformation of spatial granularity; use a circle with a radius of 6km as the initial interval to expand the research range to 24km in steps of 3km to complete the space change in magnitude.

Further, the step S7 multi-scale feature analysis specifically includes: follow the landscape pattern index selected in step S5, obtain the data of spatial granularity transformation and spatial amplitude change from step S6, and use Fragstats 4.2 software to select from patches, categories, The selected indexes are calculated at the three levels of landscape, and the multi-scale characteristics of the urban heat island landscape pattern can be analyzed by calculating the landscape pattern index at different scales.

Advantage of the present invention and beneficial effect are as follows:

Based on the landscape pattern index analysis method, the present invention uses the landscape index to represent the urban thermal landscape to analyze its various characteristics, including spatial pattern distribution and multi-scale features. Referring to the research methods of traditional landscape patterns by landscape ecology scholars in the past, we use the landscape pattern index analysis method to conduct quantitative analysis on the urban thermal landscape represented by the landscape index. When formulating the thermal landscape grading method, considering the comparative analysis among multiple research areas, it is proposed to randomly select the highest temperature in the vegetation area as the threshold for defining the heat island effect based on the traditional "mean-standard deviation" grading standard, which will be lower than the threshold The part is regarded as no heat island area, above the threshold, the part is subdivided into three or more thermal landscapes by multiples of the standard deviation of surface temperature. Secondly, when analyzing the multi-scale characteristics of the thermal landscape pattern, it is proposed to use a circle instead of a rectangle to analyze the spatial extent of the research area, which improves the current shortcomings of the spatial amplitude analysis method and makes the analysis more comprehensive. Finally, the present invention can further compare and analyze two or more research areas by calculating the three levels of landscape index of patch, type and landscape, through granularity transformation and amplitude change, so as to analyze the multi-scale characteristics of thermal landscape in different It is a scientific question whether the change trends among different types of cities are the same.

Description of drawings

Fig. 1 is a flowchart of a method for quantitative description and multi-scale feature analysis of urban thermal landscape provided by the preferred embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be described clearly and in detail below with reference to the drawings in the embodiments of the present invention. The described embodiments are only some of the embodiments of the invention.

The technical scheme that the present invention solves the problems of the technologies described above is:

The present invention is oriented to the relevant research on the urban heat island landscape pattern, and analyzes the spatial pattern and multi-scale characteristics of the thermal landscape with the help of the landscape pattern index and scale changes. Using the Landsat-8 OLI/TIRS remote sensing image as the data source, with the help of existing remote sensing image processing tools including ENVI 5.3, the original image is preprocessed including cropping, radiometric calibration, atmospheric correction and geometric correction, and further inverted The surface temperature of the study area. On the basis of the above work, the highest value of the temperature in the vegetation sample area was used as the threshold to distinguish whether there is an urban heat island effect, and the classification of the thermal landscape was completed according to the improved "mean-standard deviation" method. In the experiment, according to the geographical elements in the study area, the relevant landscape index was selected from the three levels of patch, type and landscape and calculated, and the spatial pattern of the urban thermal landscape was quantitatively analyzed according to the value of the landscape index. Finally, through the scale change and the index selected in the previous step, analyze and study the multi-scale characteristics of the regional urban thermal landscape, and further explore the change trend of this multi-scale characteristic in two or more different types of cities, and complete the analysis of the urban thermal landscape. A more comprehensive analysis of landscape spatial structure and scale effects.

Fig. 1 of the present invention has shown the method flow chart of the present invention, below with the main urban area of Chongqing and the research case of the main urban area of Chengdu as the target research area, concrete steps are as follows:

(1) Data acquisition, according to the research plan and image quality, select the Chongqing city (track number 128/39 ) and two Landsat 8 OLI_TIRS images of Chengdu (track number 129/39) on May 1, 2017. The Landsat 8 satellite is equipped with an Operational Land Imager (OLI) that provides multispectral data and a thermal infrared sensor (Thermal Infrared Sensor, TIRS) that provides multispectral data. The spatial resolution of the multispectral bands in the OLI land imager is The thermal infrared band spatial resolution of the TIRS sensor is 100m.

(2) Data preprocessing, the preprocessing is mainly based on existing software including ENVI 5.3, using other satellite images with a higher resolution than the acquired data as reference images, and selecting 30 to 40 road intersections, river bends and intersections Take the same name as the control point, and carry out the double polynomial geometric fine correction on all the bands in the Landsat 8OLI/TIRS image, including the thermal infrared band, and then use the FLAASH tool in the ENVI 5.3 software to perform atmospheric correction on each band. Obtain the data of the study area after geometric correction and atmospheric correction.

(3) Inversion of surface temperature, using the classical radiative transfer equation method to invert the surface temperature, the steps are: according to the data obtained in step S2, first divide the types of ground objects in the study area into natural surfaces, towns, and water bodies Three categories. On the basis of classification, the surface emissivity is further retrieved according to NDVI. The water body pixel is often a single water body, so the specific emissivity ε between the natural surface and the town is mainly considered. The ε at the satellite pixel scale can be calculated as follows: ε=P _V R _V ε _V +(1-P _V )R _X ε _X +dε In the formula, P _V represents the vegetation coverage, R _V is the temperature ratio of vegetation, ε _V is the specific emissivity of vegetation, R _X is the temperature ratio of bare soil or building surface, ε _X is the specific emissivity of bare soil or building surface, and dε is the thermal radiation interaction. According to the vegetation coverage, the temperature ratio of ground objects, vegetation, bare soil, and building surface temperature ratio can be further estimated. Secondly, estimate the influence of the atmosphere on the surface thermal radiation, and calculate the thermal infrared radiation brightness value received by the satellite sensor: L _λ =[εB(T _surface )+(1-ε)L _d o _wn ]τ+L _upper , where , L _λ represents the thermal infrared radiance received by the satellite sensor; ε represents the specific emissivity of the surface, which is directly brought into the result of the previous step; τ represents the transmittance of the atmosphere in the thermal infrared band; T _surface represents the true surface temperature ( K); B(T _surface ) represents the radiance of a black body with temperature T _s ; _Lupper and L _down represent the upward and downward radiance of the atmosphere respectively, and finally use Planck's formula 1.3 to obtain the surface temperature T _sruface : T _surface ＝K ₂ /[ln(1+K ₁ /B(T _surface ))], where K ₁ and K ₂ are constants, which can be obtained from the header file of the corresponding data. For the 10th band of the TIRS thermal infrared sensor, K ₁ =774.89Wm-2·sr-1·um-1, K ₂ =1321.08K.

其中，R_LST为计算出的温度阈值范围，

为研究区所选植被样本区域的最高温度，SD为地表温度标准差，n为标准差的倍数。进一步依据上述公式将研究区域划分为4种热力景观类型，依次为无热岛(Non heatisland)、弱热岛(Weak heat island)、热岛(Heat island)和强热岛(Strong heatisland)。阈值区间表示如下公式所述：(4) Select the segmentation interval as the maximum value and minimum value of the surface temperature, divide the surface temperature image into grades, and use this to reflect the type of thermal landscape. The classification formula is as follows:

where R _LST is the calculated temperature threshold range,

is the maximum temperature of the selected vegetation sample area in the study area, SD is the standard deviation of the surface temperature, and n is the multiple of the standard deviation. Further, according to the above formula, the study area is divided into four types of thermal landscapes, which are non heat island, weak heat island, heat island and strong heat island. The threshold interval is expressed as described in the following formula:

(5) Use Fragstats 4.2 to calculate the landscape pattern index. According to the relevant research results of domestic and foreign scholars, combined with the landscape characteristics of the study area, the maximum patch index (LPI), the percentage of the patch area occupied by the landscape ( PLAND), number of patches (NP), patch density (PD), landscape shape index (LSI), aggregation index (AI), sprawl index (CONTAG), Shannon diversity index (SHDI), respectively at the type level Six indexes were selected for calculation and landscape level, and the spatial pattern was analyzed according to the results of the indexes.

(6) Using the multi-scale analysis method of changing the spatial scale of the landscape pattern, starting from the original resolution of 30m in the study area, resampling the image to 300m according to the step size of 30m, and then resampling to 600m according to the step size of 60m, Finally, resample to 960m with a step size of 120m to complete the transformation of the spatial granularity; use a circle with a radius of 6km as the initial interval to expand the research range to 24km with a step of 3km to complete the change of the spatial amplitude.

(7) From the five aspects of area, density, shape, convergence and diversity, select the largest patch index (LPI), the percentage of patch area occupied by the landscape (PLAND), the number of patches (NP), and the patch density ( PD), landscape shape index (LSI), agglomeration index (AI), sprawl index (CONTAG), Shannon diversity index (SHDI), select six of them for calculation at the type level and landscape level, and at the same time in 18 The selected indexes are calculated on different granularities and seven kinds of magnitudes, and their multi-scale characteristics and changing trends between the two cities are analyzed according to the results.

The above embodiments should be understood as only for illustrating the present invention but not for limiting the protection scope of the present invention. After reading the contents of the present invention, skilled persons can make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims (7)

1. A quantitative description and multi-scale feature analysis method for urban thermodynamic landscapes is characterized by comprising the following steps:

s1, acquiring Landsat-8OLI/TIRS remote sensing satellite image data of a research area;

s2, remote sensing image data preprocessing: preprocessing the Landsat-8OLI/TIRS remote sensing image data acquired in the step S1, including radiometric calibration, atmospheric correction, geometric correction and cutting;

s3, inverting the earth surface temperature: inverting the surface temperature of the remote sensing image data preprocessed in the step S2 by adopting a classical radiation transmission equation method to obtain surface temperature data of the research area;

step S4, thermal landscape level division: obtaining the highest value of the earth surface temperature of the vegetation area by random sampling, and establishing a thermodynamic landscape grading standard by adopting a mean-standard deviation method;

step S5, selecting the landscape pattern index: according to the geographic environment and the research requirement of the research area, required indexes are selected from three levels of plaques, categories and landscapes, the thermodynamic landscape data obtained in the step S4 are substituted into the calculation to obtain a result, and the spatial distribution characteristics of the urban heat island are quantitatively analyzed according to the result;

step S6, scale change: performing granularity conversion on the thermal landscape data acquired in the step S4 in terms of spatial granularity, and performing amplitude change on the thermal landscape data acquired in the step S4 in terms of spatial amplitude;

s7, performing multi-scale feature analysis, namely calculating index results of data with different granularities and amplitudes at three levels of patches, categories and landscapes from the spatial granularity and the spatial amplitude according to the landscape pattern index selected in the step S5 and the scale change performed in the step S6, and obtaining the multi-scale features of the urban heat island landscape according to the response relation of the urban heat island landscape analysis index represented by the landscape index along with the change of the spatial scale;

the step S4 of obtaining the highest value of the earth surface temperature of the vegetation area by utilizing random sampling, and then formulating the thermodynamic landscape grading standard by adopting a mean-standard deviation method specifically comprises the following steps:

for the classification of the single-band gray level image, selecting a segmentation interval as a maximum value and a minimum value of the earth surface temperature, classifying the earth surface temperature image, and reflecting the type of the thermal landscape according to the classification formula as follows:

wherein R is _LST For a calculated temperature threshold range, ->

The method comprises the following steps of (1) dividing a research area into 4 thermal landscape types according to a classification formula, wherein the maximum temperature of a vegetation sample area selected for the research area is SD (standard deviation of earth surface temperature), n is a multiple of the standard deviation, the thermal landscape types are sequentially a Non Heat island, a Weak Heat island, a week Heat island, a Heat island and a Strong Heat island, and the threshold value is expressed by the following formula:

2. the method for quantitative description and multi-scale feature analysis of urban thermal landscape according to claim 1, wherein the step S2 of preprocessing the remote sensing image data specifically comprises: by means of existing software including ENVI5.3, other high-resolution satellite images are taken as reference images, 30-40 same-name feature points at road intersections, river turns and intersections are taken as control points, all wave bands including thermal infrared wave bands in the Landsat 8OLI/TIRS images are subjected to quadratic polynomial geometric fine correction, then radiation calibration and atmospheric correction are carried out on all the wave bands by using a radiation calibration tool and a FLAASH tool which are carried in the ENVI5.3 software, and preprocessed data are obtained and cut according to requirements so as to obtain final research area data.

3. The method for quantitative description and multi-scale feature analysis of urban thermal landscape according to claim 2, wherein said step S3 is for inversion of surface temperature by using classical radiative transfer equation method, comprising the steps of: according to the preprocessing data of the research area obtained in the step S2, firstly, the ground object types of the research area are divided into three categories of a natural surface, a town and a water body, the ground surface emissivity is inverted according to the normalized vegetation index NDVI, the water body pixel is often a single water body, and therefore the ground surface emissivity epsilon of the natural surface and the town is considered, and the epsilon is calculated according to the formula: ε = P _V R _V ε _V +(1-P _V )R _X ε _X + d ε formula wherein P _V Indicating vegetation coverage, R _V Is the temperature ratio of vegetation, ε _V Is the specific radiance of vegetation, R _X Is the temperature ratio of bare soil or building surface, epsilon _X The specific radiance of bare soil or the surface of a building is adopted, d epsilon is the thermal radiation interaction, and the temperature ratio of the ground and the living things, the temperature ratio of vegetation, bare soil and the surface of the building are estimated according to the vegetation coverage; secondly, estimating the influence of the atmosphere on the earth surface thermal radiation, and calculating the thermal infrared radiation brightness value received by the satellite sensor: l is _λ ＝[εB(T _surface )+(1-ε)L _down ]τ+L _upper In the formula, L _λ Indicating the brightness of the thermal infrared radiation received by the satellite sensor; epsilon represents the surface emissivity; τ represents the transmittance of the atmosphere in the thermal infrared band; t is _surface Representing the true surface temperature (K); b (T) _surface ) Denotes a temperature of T _s Black body radiation brightness of (d); l is _upper And L _down Respectively representing the upward and downward radiation brightness of the atmosphere, and finally obtaining the surface temperature T by utilizing the Planck formula 1.3 _sruface ：T _surface ＝K ₂ /[ln(1+K ₁ /B(T _surface ))]In the formula K ₁ ，K ₂ All of which are constants, and can be obtained from the header file of the corresponding data.

4. The method for quantitative description and multi-scale feature analysis of urban thermodynamic landscape according to claim 3, wherein the step S5 of selecting a landscape pattern index specifically comprises: fragstats 4.2 is used for calculating a landscape pattern index, required indexes are selected from three levels of patches, categories and landscapes by combining the landscape characteristics of a research area for calculation, and the spatial distribution characteristics of the urban thermal landscape are analyzed from the aspects of area, density, shape, clustering property and diversity according to the index result.

5. The method of claim 4, wherein the maximum plaque index LPI, percentage of plaque area occupied by landscape PLAND, number of plaques NP, plaque density PD, landscape shape index LSI, aggregation index AI, vintage index CONTAG, and Shannon diversity index SHDI are selected from the five aspects of area, density, shape, clustering, and diversity.

6. The method for quantitative description and multi-scale feature analysis of urban thermodynamic landscape according to claim 4, wherein the step S6 is to perform granularity transformation on the thermodynamic landscape data obtained in the step S4 from a spatial granularity and perform amplitude change on the thermodynamic landscape data from a spatial amplitude, and specifically comprises:

determining the spatial amplitude of the research by the total area of the research area, sequentially resampling the image to 300m by adopting a landscape pattern multi-scale analysis method for changing the spatial scale with the original resolution of 30m as a starting point in the research area and the step length of 30m, then resampling to 600m by adopting the step length of 60m, and finally resampling to 960m by adopting the step length of 120m to complete the transformation of spatial granularity; and (3) sequentially expanding the research range to 24km by taking a circle with the radius of 6km as an initial interval according to the step length of 3km to finish the change of the space amplitude.

7. The method for quantitative description and multi-scale feature analysis of urban thermodynamic landscape according to claim 6, wherein the step S7 of multi-scale feature analysis specifically comprises: and (3) following the landscape pattern index selected in the step (S5), acquiring data for completing space granularity transformation and space amplitude change from the step (S6), calculating the selected index from three levels of patches, categories and landscapes in Fragstats 4.2 software, and performing multi-scale feature analysis on the urban heat island landscape pattern by calculating the landscape pattern indexes under different scales.

Images