You must initiate ENVI batch in order for IDL to recognize ENVI routines such as the radiometric calibration task. e = ENVI(/HEADLESS) and e.Close

Another question to consider would be which version of ENVI you are working with. The majority of ENVITasks made their way into ENVI at version 5.2, so if you have a previous version of ENVI, I would expect an error like that since ENVI would not know what to do with it. For reference, here is my code for radiometrically calibrating Landsat data using ENVIRadiometricCalibrationTask in ENVI 5.2 in headless mode (just make sure to edit my file paths): PRO Landsat_RadCal_Batch compile_opt idl2 ; Start the application e = ENVI(/HEADLESS) ; A Landsat dataset consists of one or more TIFF files, ; with an associated metadata file (*MTL.txt). Open the ; metadata file to automatically read the gains and offsets ; necessary to perform radiometric calibration. ; Create a list of input files to perform radiometric calibration filelist = File_Search('C:\Data\Landsat\SierraReservoirs', '*MTL.txt') ; Get the radiometric calibration task from the catalog of ENVI tasks. Task = ENVITask('RadiometricCalibration') ; Define the output data type and calibration type Task.Output_Data_Type = 'Float' Task.Calibration_Type = 'Top-of-Atmosphere Reflectance' ; Define output location out_dir = 'C:\Data\Landsat\SierraReservoirs\RadCal' ; Loop over every file. ; Open the raster, set parameters, and execute the task. foreach file, filelist do begin output_file = out_dir + path_sep() + 'radio_' + file_basename(file) raster = e.OpenRaster(file) task.input_raster = raster[0] task.output_raster_uri = output_file task.Execute endforeach ; Get the data collection DataColl = e.Data ; Add the output to the data collection DataColl.Add, Task.Output_Raster ; Close the ENVI session e.Close END

Thanks for your answers I think I find the problem. I have ENVI 5.0.2 and I think this version does not support ''Radiometric Calibration'' as a task. So I must generate my own function based on 'TMCAL_DOIT' routine. Unfortunately I find that this routine just works with old-style metadata file of LandSat. Is there any suggestion for my problem? Best Regards,

Hello, The TMCAL_DOIT routine was updated to work with the new Landsat metadata files. In your code, open the associated metadata file (_MTL.txt), then set the /USE_METADATA keyword. Please see http://www.exelisvis.com/docs/TMCAL_DOIT.html for details. the TMCAL_DOIT help topic in the ENVI Classic help may also have a code example.

To>> joe peters Hi, I also have problem with doing landsat8 calibration, and I am newbie using envi/idl :D I am using Envi 5.1 and IDL 8.3 I have tried code, called ENVIRadiometricCalibrationTask provided by Exelis VIS > and I also have tried your posting code. After compiled and run the program, the execution halted with this error message "Input value is of type STRING, which cannot be reliably cast to numeric type INT" the error message refer to below line ; Define the output data type and calibration type Task.Output_Data_Type = 'String' Task.Calibration_Type = 'Top-of-Atmosphere Reflectance' I could not find how to convert string to integer type, could you please help me? By the way, does your posted program run well in your IDL? thanks ^_^

Why are you setting the output data type of your raster to a string? Task.Output_Data_Type = 'String' It is a raster, so you will want the pixels to be represented by integers, floats, etc. The data type options are: OUTPUT_DATA_TYPE (optional) Specify a string indicating the output data type: Int: 16-bit integer Float: Floating-point (default) Double: Double-precision floating point UInt: 16-bit unsigned integer In the example code they use a string to specify the data type: Task.Output_Data_Type = 'Double'

Task.Output_Data_Type = 'String' Sorry, I have mistyped above line but actually I have tried the code with output data type set as float Task.Output_Data_Type = 'Float' and the error message: " Input value is of type STRING, which cannot be reliably cast to numeric type INT"

This appears to work correctly in ENVI 5.2 so it may have been a bug in 5.1. Can you upgrade to 5.2?