I am doing UHI change and my study area is covered by two Landsat images (two paths and one row). So is combine the two image will give the accurate temperature or not. Do you please have a suggestion or have you came across a study with same issue?
Hi Ali
If the images belong to the same day then there is no problem otherwise there will be difference in the temperature (for different day) and then it will not be meaningful. So try to collect images for the same days and then you cam mosaic them and use for your study.
All the best.
If you combine Landsat images you have to be aware, that they may have different timestamps - therefore a atmospheric correction has to be applied.
UHI is very time specific and two landsat images with time lag cannot be effectively used for UHI (LST). However the temperature anomalies should not be very different, it can be the absolute value that is going to be different.
Try to find out the scenes with least time gap, it won't introduce much error. Even if there is error it can be easily corrected by adding the constant error value.
Another option is to try other data sets which provide LST like MODIS, (provided the coarse resolution serve your purpose).
Yes MODIS should give UHI as LST expression but have not seen any outputs and publications.
Since Landsat images filmed on the same day (two paths and one row), then there is no problem.
Atmospheric correction must be made, and then create a mosaic of adjacent images and conversion of DN at-sensor radiance and conversion of at-sensor radiance to the temperature to get the real heat of the surfer
I guess this Landsat image can have a significant offset in temperature. But the atmospheric correction should help.
Many precautions are needed before one can combine two TIR scenes. Firstly, do they belong to the same date and time of acquisition? If so, you can combine after performing the atmospheric correction and Emissivity-Temperature separation. If they are not, obviously, you need to estimate the surface kinetic temperatures separately and combine the derived thematic details.
Agree Rama actually emissivity is quite unlikely to be different. So the entire temp bias will be due to atmospheric effect.
Yes Dr.Saif Uddin, atmospheric effects are significant when the emissivity contrast is less.(Since it is Landsat data -with coarse spatial resolution - the changes in roof emissivities may be ironed out). But if the data belongs to two different periods, the irradiating heat flux can also cause significant changes in brightness temperature between data sets.
- Badges
- Science topic
- Similar topics
- Geoscience
- Geoinformatics