Try some print statements to see what is returned when you do File_search and what the value of 'ProductCount' is. Do these return valid values (does it find files)? You are passing a string ('file') to ENVI_OPEN_DATA_FILE but it looks like file is a variable. Do a print on 'file' and see what its value is. Is it a file name?

it is strange. today I try the pro again. except the ENVI_DOIT, 'TMCAL_DOIT' looks like not run, the other are work well, and the fid =10 ,any one can tell why the ENVI_DOIT, 'TMCAL_DOIT' function does not run?

I can think of a lot of reasons it wouldn't run. But they would all have different behavior. Can you tell us exactly what you are seeing when it doesn't run? Do you get an error message? Peg

actually, as a vegetation bird, I write this pro file which is modified from the discussion "http://www.exelisvis.com/...posts.aspx",and I may have not enough exprience to finish it. hoep anyone give me help./ the result is show like this : "H:\LT51430302010232IKR00\L5143030_03020100820_MTL.txt 10 10 sun elevation=53.6730895 sun_azimuth=140.908264 IDL> "

It's hard to understand exactly what is happening from your description, but it looks to me like your code ran successfully. You show the program returning to an IDL prompt, and no error message, which is the behavior I would expect if everything ran successfully. Have you checked to see whether the output image you specified in your code was created, and if so, whether it looks good? - Peg

For Landsat 7 (Landsat 5 too), code will look something like this: [QUOTE]Code example PRO LANDSAT_CALIBRATION_BATCHMODE COMPILE_OPT IDL2 ENVI_BATCH_INIT InputPath = 'c:\LANDSAT\' ListFiles = FILE_SEARCH(InputPath, COUNT = ProductCount, '*MTL.txt', /FOLD_CASE, /TEST_READ, /FULLY_QUALIFY_PATH) IF(ProductCount LE 0) THEN BEGIN Print, 'There are no valid landsat metadata to be processed.' RETURN ENVI_BATCH_EXIT ENDIF FOR i=0, ProductCount-1 DO BEGIN InputFile = ListFiles[i] Filename = FILE_BASENAME(InputFile, '_MTL.txt', /FOLD_CASE) Envi_open_data_file, InputFile, /landsat_metadata fids = envi_get_file_ids() IF (fids[0] EQ -1) THEN BEGIN Print, 'There are no valid landsat metadata to be processed.' RETURN ENVI_BATCH_EXIT ENDIF FOR j = 0, n_elements(fids)-1 DO BEGIN envi_file_query, fids[j], fname = fname, nb=nb IF (nb EQ 6) OR (nb EQ 4) THEN BEGIN image_fid=fids[j] envi_file_query, image_fid, fname = fname_image, dims=dims_image, nb=nb_image print, fname_image , ' nb: ', nb_image, ' fid: ', image_fid ENDIF ENDFOR pos=lindgen(nb_image) ; Metadata parser n = FILE_LINES(InputFile) metadata = STRARR(1,n) OPENR, unit, InputFile, /GET_LUN READF, unit, metadata FREE_LUN, unit ; Find position of sun_angle in metadata and read value ind=WHERE(STRPOS(metadata,'SUN_ELEVATION') ge 0) IF (ind NE [-1]) THEN BEGIN searching_line = metadata[ind] searching_line_data=strsplit(strcompress(searching_line, /remove_all),'=', /extract) sun_elevation=float(searching_line_data[1]) sun_angel=90.0-sun_elevation ENDIF ; Find position of date in metadata and read value ind=WHERE(STRPOS(metadata,'ACQUISITION_DATE') ge 0) IF (ind NE [-1]) THEN BEGIN searching_line = metadata[ind] searching_line_data=strsplit(strcompress(searching_line, /remove_all),'=', /extract) date_line=searching_line_data[1] date_array=fix(strsplit(strcompress(date_line, /remove_all),'-', /extract)) date=intarr(3) date[0]=date_array[1] date[1]=date_array[2] date[2]=date_array[0] print, date ENDIF ; Determine Satellite IF ((STRPOS(Filename,'L7')) NE -1) THEN BEGIN sat=0 bands_present=[0,1,2,3,4,7] ENDIF IF ((STRPOS(Filename,'L5')) NE -1) THEN BEGIN sat=1 bands_present=[0,1,2,3,4,6] ENDIF PRINT, 'sun elevation=', sun_elevation PRINT, 'sat_ind=', sat gain=fltarr(nb_image) bias=fltarr(nb_image) min_max_rad=WHERE(STRPOS(metadata,'MIN_MAX_RADIANCE') ge 0) lmax_min=fltarr(min_max_rad[1]-min_max_rad[0]-1) searching_line = metadata[min_max_rad[1]+ 2] searching_line_data=strsplit(strcompress(searching_line, /remove_all),'=', /extract) qcalmax=fix(searching_line_data[1]) searching_line = metadata[min_max_rad[1]+ 3] searching_line_data=strsplit(strcompress(searching_line, /remove_all),'=', /extract) qcalmin=fix(searching_line_data[1]) FOR b=0, n_elements(lmax_min)-1 DO BEGIN searching_line = metadata[min_max_rad[0]+b+1] searching_line_data=strsplit(strcompress(searching_line, /remove_all),'=', /extract) lmax_min[b]=float(searching_line_data[1]) ENDFOR IF (sat EQ 0) OR (sat EQ 1) THEN BEGIN lmax_min=reform(lmax_min, 2, n_elements(lmax_min)/2, /overwrite) gain[0]=(lmax_min[0,0]-lmax_min[1,0])/(qcalmax-qcalmin) bias[0]=lmax_min[1,0]-((lmax_min[0,0]-lmax_min[1,0])/(qcalmax-qcalmin))*qcalmin gain[1]=(lmax_min[0,1]-lmax_min[1,1])/(qcalmax-qcalmin) bias[1]=lmax_min[1,1]-((lmax_min[0,1]-lmax_min[1,1])/(qcalmax-qcalmin))*qcalmin gain[2]=(lmax_min[0,2]-lmax_min[1,2])/(qcalmax-qcalmin) bias[2]=lmax_min[1,2]-((lmax_min[0,2]-lmax_min[1,2])/(qcalmax-qcalmin))*qcalmin gain[3]=(lmax_min[0,3]-lmax_min[1,3])/(qcalmax-qcalmin) bias[3]=lmax_min[1,3]-((lmax_min[0,3]-lmax_min[1,3])/(qcalmax-qcalmin))*qcalmin gain[4]=(lmax_min[0,4]-lmax_min[1,4])/(qcalmax-qcalmin) bias[4]=lmax_min[1,4]-((lmax_min[0,4]-lmax_min[1,4])/(qcalmax-qcalmin))*qcalmin gain[5]=(lmax_min[0,6]-lmax_min[1,6])/(qcalmax-qcalmin) bias[5]=lmax_min[1,6]-((lmax_min[0,6]-lmax_min[1,6])/(qcalmax-qcalmin))*qcalmin ENDIF print, qcalmax print, gain print, bias ENVI_DOIT, 'TMCAL_DOIT', fid=image_fid, bands_present=bands_present, pos=pos, dims=dims_image, sat=sat, $ cal_type=1, sun_angle=sun_angel, out_name=InputPath+Filename, r_fid=r_fid, date=date, $ gain=gain, bias=bias FOR index=0, n_elements(fids)-1 DO BEGIN envi_file_mng, /remove, id=fids[index] ENDFOR ENDFOR envi_batch_exit END [/QUOTE] I'm not sure how well TMCAL_DOIT works, but the description in the help causes serious though . [QUOTE]ENVI help result=Input * GAIN + BIAS gain = (lmax - lmin) / 255 bias = lmin[/QUOTE] This is correct only for QCALMAX=0. Alex

TMCAL_DOIT has been re-written to pass the metadata file directly to the routine so that you no longer need to parse the metadata file. It can now be called using something as simple as: envi_doit, 'tmcal_doit', fid=fid, pos=pos, dims=dims, cal_type=1, out_name='tm_refl.dat', /use_metadata Note that you don’t need to pass in any gain/bias/date/sat/sun_angle or bands_present and you can use pos if you want to calibrate a spectral subset. We have a patch that can be applied to ENVI 4.8 that will allow for you to use the new routine. If you are interested, please contact Technical Support directly.

If you have multiple "MTL.txt" files the code provided by Alex will not work well. If you replace (almost at the end): FOR index=0, n_elements(fids)-1 DO BEGIN envi_file_mng, /remove, id=fids[index] ENDFOR with: fids=envi_get_file_ids() FOR index=0, n_elements(fids)-1 DO BEGIN envi_file_mng, /remove, id=fids[index] ENDFOR that will solve the problem (works for me). Otherwise, youll be always using the first file. Regardless this small "bug", the code works perfectly (thank you Alex for your code, was exactly what I was looking for!) Im also not sure if you have to use the sun elevation raw value, or you have to convert it to sun angle (like Alexs code does)... Im always confused about it. It seems to me that when you do it manually (through the ENVIs menu), ENVI uses the sun elevation value; Am I correct? Álvaro.

Yes, the SUN_ANGLE keyword is for the solar elevation angle, which is the angle of the sun above the horizon, extending from 0 degrees to 180 degrees. I think 0 to 90 to 180 degrees is when it's over the western horizon. So, 90 degrees is straight overhead. - Peg

The code works well in TM5 visible bands calibration. But when I calibration the TM5 thermal band to brightness temperature, I have set band_present=[5], the value is always between 0-1, not in Kelvin. I hope you give my any suggestion, thanks very much.

The documentation states: Set BANDS_PRESENT=[0] to process Landsat-5 TM band 6 thermal data. This band is handled separately in the Available Bands List and has its own file ID (FID). http://www.exelisvis.com/docs/TMCAL_D...