Hello Hamouda,

I am also determining ET from remote sensing (Landsat 7/8 and airborne data). In my opinion,the easiest way to determine albedo, LST, emissivity NDVI... is by using "band math". There you should include the formulae of your parameter and the software will give you a raster of your variable.

Is that what you are asking for? I hope it can help you.

As mentioned above you can can use band math and then save your data as TIFF files which you can import in ILWIS

Thank you Dr.Cuesta and Dr.El Sharkawy. Do you have like a tutorial for that?

Attached some reference may help

http://www.yale.edu/ceo/Documentation/Landsat_DN_to_Kelvin.pdf

https://www.itc.nl/ilwis/documentation/version3.asp

Thank you Dr.El-Sharkawy

I  use ERDAS imagine model builder to execute ET models 

In my opinion ERDAS is useful to highlight the relationship between variables, but the final models are not user-friendly (lot of lines, boxes, etc...). I have generated toolboxes in ArcGIS in order to the users select their inputs and run the models. They do not see the relationship between variables but it is more user friendly instead. 

I think the ArcGIS is not  like ERDAS or ENVI as Image Processing software, so the results will not accurate 

I have compared both ERDAS, ENVI and ArcGIS results and I can ensure that the results are the same when pplying energy balance models

Yes, but at the end, you need to compare the results of solar radiation and ET based on LULC  classes, for that you need to use ENVI which have accurate classification algorithms.