readme.txt

---------------------------------------------------------
GEWEX Asian Monsoon Experiment-Tropics (GAME-T) dataset

******* Daily photograph from EGAT tower ************
*******   Version 1.00   ****************************

(C) 2002/3/4 Kenlo Nishida, nishida@bres.tsukuba.ac.jp
---------------------------------------------------------

Camera location: EGAT Tower 60m terrace. Southward direction.

Camera: Minolta alpha 707si with semi-fisheye lense

Interval: 24 hours (automatic; film was exchanged every month by Mr. Wichai @ RID)

Time of capture: approximately noon (local time)

period: 1999/5/12 - 2000/11/15  (in this version)

missing period: 1999/8/11-8/13, 2000/6/16  
    (instead, surrogated and filled with pictures of nearest date)

processing:

   1. photograph (negative film) development
       - In a photoshop in Pitsanulok 
       - Print: A: 12.6cm*8.9cm paper (most of the photo)
                B: 15.1cm*9.9cm paper (99/6/12-7/11)
        
   2. digitized by a scanner from prints
       - Scanner: Cannon CanoScan N1220U
       - for case A, 4.7inch*3.4inch, 200dpi.
         for case B, 5.45inch*3.95inch, 180dpi.
         
   3. view position adjustment
       - pixel pattern matching (cross-correlation) on IDL
         template image is "EGAT99-0138.JPG"
         
   4. brightness adjustment
       1 Look at the lower half of the image (land)
       2 Calculate the average brightness of that area
         for both template (b1) and target (b2)
       3 ratio=(1 + b1/b2)/2  ..... half adjustment
       4 multiply this ratio on all the pixels.
       
Files:
   - all the data are stored in JPEG format.
   - file names:
      .... "EGATyy-0jjjc.JPG"
   (note: yy is 2-digit year, jjj is 3-digit julian-day, c means "corrected" by process 3 and 4)
   
Ancillary files:
   - a.pro .... IDL program for adjustment (process 3 and 4). (C) K. Nishida
   
Note: 
   - Brightness of the original images depend on development (because
     they are separately developped, both negative film and paper prints).
     Therefore, brightness abruptly change from film roll to film roll.
     The following is the division of the film rolls:
     
     99/5/12-6/11 (one roll)
     99/6/12-7/11
     99/7/12-8/10
     99/8/14-9/18
     99/9/19-10/20
     99/10/21-11/20
     99/11/21-12/21
     99/12/21-00/1/17
     00/1/18-2/18
     00/2/19-3/18
     00/3/19-4/19
     00/4/20-5/17
     00/5/18-6/14
     00/6/15-7/14
     00/7/15-8/14
     00/8/15-9/15
     00/9/17-10/16
     00/10/17-11/15