Land Surface Temperature L8

[diegogeofm44]

Hello, I'm trying to retrieve LST from L8, and I've been reading about SWA to get LST, and I need the band 10 and 11 for this case, but the imagery provided in landsat 8 L2C2 doesn't show the band 11, which is very important to do the algorithm. Does anyone know where I can find the band 11, or if it has a different name on the products.

I'm also deciding of using the method from U. Advan and G. Jovanovska (2016) which uses only band 10 and is quicker, or apply the SWA purposed for O. Rozenstein, et. al. (2014) which uses bands 10 and 11

Hope someone can help me, :)

[LP DAAC - dgolon]

HI @diegogeofm44 Thanks for writing in. Band 11 is not used in the generation of Landsat Level-2 Surface Temperature or Surface Reflectance products. Therefore, it is not included in those downloads. This is similar to band 8 (Panchromatic). If you are working with the Level-2 data you do not need to calculate LST, as it is already presented as Surface Temperature / Surface Reflectance. While we recommend working with Level 2 data, if for some reason you absolutely need to calculate the LST yourself, then we recommend using the Level-1 data. More information on these data products can be found here: https://www.usgs.gov/landsat-missions/landsat-collection-2-level-2-science-products Thanks -- Danielle

[diegogeofm44]

Makes a lot of sense because I've seen a few videos and read articles that uses Level 1 data, also read what the product guide says about that you're telling me, it says that "Landsat 8-9 TIRS-only data (LT08/LT09) cannot be processed to TOA Brightness Temperature or Surface Temperature".
But still got me confused, because the data presented in band 10 is presented in digital numbers (DN), right.
So if I'm well understanding, I still should do the procces to have that DN to Kelvin and then Celsius, so I can get a temperature map.
And this procces will be TOA Radiance to BT, then calculate NDVI and PV to have emissivity (E), and finally have LST. Right?

[LP DAAC - dgolon]

Hello @diegogeofm44 Here is what I heard back from the team:

1) Level-1 data is essentially TOA data, you need to do a couple calculations in order to get Brightness Temperature, Reflectance, Radiance, etc. However, we cannot calculate Surface temp from TIRS only because there is information needed in the visual bands (OLI) in order to match up Landsat data to the Auxiliary data needed to calculate Surface Temperature. We collect nighttime data (both OLI and TIRS) and most of it cannot be processed to Level-2 (Surface Temp, etc.) because even though we "have" OLI bands they are all black and we cannot find reference points.

2) Landsat is stored / distributed with DN values because of storage size. It is not related to TOA, Brightness Temperature, Etc. You need to apply a scale factor (listed on the page linked earlier under the ST category) to convert from DN to Kelvin. If you wish to convert Celsius you can do that step as well. However, applying the scale factor to level-2 data will give you LST values with no additional processing needed.

What you are describing in your posts sounds like what would have needed to be done with Collection 1 data. However, now that Collection 2 data is available that is not needed to be done with the Collection 2 data. Part of the reason we released Collection 2, and specifically Level-2 products, was to a) give more accurate values (including Aux data has made it more accurate) and b) make it easier for users.

Please let us know if you have any additional questions.

[diegogeofm44]

Can't thank you enough for making my mind clear, now I totally understand everything that you're saying.
Really appreciate your help, have a good one.

Best regards from Mexico :)

[LP DAAC - dgolon]

Glad to hear it! Please let us know if you have any future questions. Thanks and Happy Holidays!