Howdy! I've been trying to calculate the land surface temperature of Ghana by following exactly this post: https://www.gislounge.com/how-to-use-arcgis-pro-to-map-urban-heat-islands. I have used Band 4, Band 5 and Band 10 I downloaded from USGS. However, after using the following formulas and creating my LST outcome layer, I get weird extreme temperatures (61 to 91 degrees) and I don't understand why:
My formulas and values are:
1) Top of Atmosphere (TOA) = 0.00038*”B10.TIF”+0.1
Radiance Multiplier = RADIANCE_MULT_BAND_10 = 3.8000E-04 (so 0.00038)
Radiance Add = and RADIANCE_ADD_BAND_10 = 0.10000, both from the metadata
2) Brightness Temperature (BT) = (1329.2405/Ln((799.0284/"TOA")+1))-273.15
K1_CONSTANT_BAND_10 = 799.0284
K2_CONSTANT_BAND_10 = 1329.2405, both from the metadata
3. NDVI = Float (band5-band4) / (band 5+band 4)
4. Proportional Vegetation (PV) = Square((“NDVI”+0.121068)/(0.459455+0.121068))
NDVI min is -0.121068
NDVI max is 0.459455, both from NDVI outcome layer
5. Error Correction / Emissivity (E) = 0.004* ”PV”+0.986
6. Land Surface Temperature (LST): (“BT”/(1+(0.00115*”BT”/1.4388)*Ln(“E”)))
By looking at the formulas above, could someone help poor me identify any issues? 🙁
I didn`t get any operation errors while running the different steps/formulas in the raster calculator, and have tried different things to correct that, including extract by mask tool as in this video where the guy got the same error: https://www.youtube.com/watch?v=EzavTBq-1wM&ab_channel=VikasGhadamode, but it didn't work for me.
Considerations: In the LST formula, I`m not sure if using 0.00115 is right, since I`m using Band 10. But I also tried to use 0.00106 as per below, but it wouldn't work either. Maybe there is a value wrong somewhere and I can't see it...
Band 10 TIRS 1 ---- 10.6 – 11.19
Band 11 TIRS 2 ---- 11.5 – 12.51
In case anyone is feeling super nice and wanna help me solve this mystery 😅, here's the link to download the files I'm using: https://we.tl/t-Pum2lyBqWU
Thank you so much - forever greatful 😎
Roger
