foreword
The software processing technology of the remote sensing image processing platform (The Environment for Visualizing Images) covers image data input/output, image calibration, image enhancement, correction, orthorectification, mosaic, data fusion and various transformations, information extraction, image classification , Knowledge-based decision tree classification, integration with GIS, DEM and terrain information extraction, radar data processing, 3D display analysis.
<ENVI Land Surface Temperature Retrieval——Single Window Algorithm>
【principle】
The principle of temperature inversion: All objects with a temperature higher than absolute zero can produce thermal radiation. The higher the temperature, the greater the total energy radiated.
Single-window algorithm: Invented by Qin Zhihao, Institute of Agricultural Resources and Agricultural Regional Planning, Chinese Academy of Agricultural Sciences. The principle is not overly explored, the point is that the method is used by me.
【Data Sources】
Geospatial Data Cloud http://www.gscloud.cn/
NASA official website https://www.nasa.gov/
Landsat website https://www.usgs.gov/core-science-systems/nli/landsat
01
data preprocessing
1.1 Data Acquisition
The geospatial data cloud selected the landsat8 image on July 10, 2017. The imaging effect is ideal. The orbit number is 123/33. The image quality in the study area is good. The spatial resolution is 30m.
1.2 Radiance temperature
1.2.1 Radiation Calibration and Brightness of Thermal Infrared Band
First import the image, perform radiometric calibration, and select Thermal as the band.
1.2.2 Calculation of radiance temperature
This formula needs to be used, Ti is the radiance temperature (radiation brightness temperature), and Lλ is the radiance obtained above (that is, the picture with a value between about 5-11).
K1=774.89, K2=1321.08.
Then we can start our band operation.
Right-click to give color grading, and adjust it to be concise and beautiful according to the temperature range. (All calculation graphs can be color-graded, and no grading will not affect subsequent operations)
1.3 Surface emissivity
Get vegetation coverage images. The specific process is to do radiometric calibration on the three multi-spectral remote sensing images respectively, then make a mosaic, and then calculate ndvi, and then use ndvi to calculate the vegetation coverage.
1.3.1 Multispectral radiometric calibration
Radio radiation analysis is selected as MultiSpectral, and then click OK. Change output interleave to BIL, then apply flaash settings to change scale factor to 0.10, and set the output layer name in output filename.
1.3.2 NDVI calculation
The maps obtained by multispectral radiometric calibration are processed by NDVI.
In order to ensure that the ndvi value is accurate (between -1 and 1), enter the statistics-spatial statistics verification. The NDVI values -0.325490 and 0.584314 for cumulative percentages at 5% and 95% were recorded.
1.3.3 Vegetation coverage calculation
Method 1: NDVI numerical calculation
*(b1 lt NDVI MIN)0+(b1 gt NDVI MAX)1+(b1 ge NDVI MIN and b1 le NDVI MAX)((b1-NDVI MIN)/ (NDVI MAX-NDVI MIN)) Get the ndvi in the previous step Two values (-0.325490 and 0.584314) replace NDVI MIN and NDVI MAX respectively.
Method 2: Band Calculation
Using the formula of (near-infrared band-red band)/(near-infrared band+red band), use the NDVI formula under transform in envi to combine bands when needed (that is, files combined with multiple bands, if each band is separated) cannot be used) NDVI=(float(b4)-float(b3))/(float(b4)+float(b3))
1.3.4 Calculation of surface emissivity
After obtaining the vegetation coverage map, the surface emissivity can be calculated. Use this formula: 0.004*b1+0.986
02
Surface temperature inversion
2.1 Intermediate parameter calculation
C is obtained by multiplying the surface specific emissivity and the atmospheric transmittance. C=b1*0.73
Atmospheric transmittance calculation: Enter the image forming time and the longitude and latitude of the center on the NASA official website atmcorr.gsfc.nasa.gov/, and the atmospheric transmittance will be provided.
The intermediate parameter D is also calculated by band
D=*(1-0.73)*(1+(1-b1)0.73)
2.2 LST inversion of land surface temperature
Atmospheric average effect temperature is calculated by querying historical weather as near-surface temperature. Because my main picture was taken at 2:53 in the morning, it was close to the lowest temperature of the day. The lowest temperature of the day was found to be 25 degrees Celsius, and here it is converted into a Kelvin temperature of 298.15. So T0=298.15,
Then Ta=2973.63353.
Finally, subtract 273.15 to convert to degrees Celsius
(-62.735657(1-b1-b2)+b3(0.434036*(1-b1-b2)+b1+b2)+b2Ta)/b1** (a=-62.735657
when the temperature is between 0-70°C , b=0.434036, this is the empirical coefficient)
The above formula b1 is C, b2 is D, b3 is the radiation brightness temperature
It needs to be calculated that Ta is the average temperature of the atmosphere, and T0 is the temperature near the surface.
Above, the surface temperature inversion of the downtown area of Tianjin has been completed.
Kind tips
The surface temperature retrieval generally uses the single-window algorithm and split-window algorithm. This article demonstrates the "honest man" step-by-step operation method; if students are interested in this, this official account will bring you other easier methods next time. For example:
> " Armstrong " Act
As the first astronaut to set foot on the moon, Armstrong said, "One small step for me, one giant leap for mankind." Also in the surface inversion method, the ENVI expansion tool can be used to obtain the atmospheric section coefficient simply and quickly so that calculate.
> "Steal half a day's leisure" method
The team of Mr. Ren Huazhong from the Institute of Remote Sensing and Geographic Information, Peking University released a software called "LSTfromL8_PKU", which can directly use the vegetation coverage method and land surface classification method to perform temperature inversion through different image representative bands of Landsad.
Source of this article: Resilient Urban Planning