# locations.table # # Note: each line which contains a hash sign (#) on the very beginning # is considered to be a comment. Also blank lines are skipped. # # The columns to provide are: # 1. beginning date (in non-american notation: dd.mm.yy) # 2. beginning time (in 24 hours notation: hh:mm:ss) # 3. ending date # 4. ending time # 5. time base: UTC, MEZ, MESZ, ADT or number of hours difference to # UTC (this is local time minus UTC) # 6. measuring height above ground (meters) or, if you measured # measuring height above displacement height, then give this value # here and set the vegetation height and displacement height to NA # 7. vegetation height above ground (meters) or NA # 8. displacement height (meters) or NA # 9. wind direction offset (angle between geogr. north and sonic north) # 10. number of analog channels; this must be the same number as in the # header of the raw data file, if you do not want to end up # having problems later somewhere. # 11. and more: depending on the settings in column 10. Starting # with column 11 the number of columns defined in 10 will # read as time lags in number of raw records (so this value # depends on the sampling rate!). Note: if you specify NU (not used) # in the lag column(s), this is equivalent to an exclude statement # in the calibration table when using the POLLUMET concept. # The next column is the prefix of the dataset (or NA). # The remaining columns are read in and concatenated to obtain # a string describing this location. # # If you choose the USEARCSSCONCEPT (see calibration.h), convertall will also # check whether the LOCATIONs match between the calibration table specified # by DEFAULTCALIBTAB in convertall.h, since it could be unclear from the # data filename, from which location the data are. # In the USEPOLLUMETCONCEPT configuration it is assumed that all data file # names start with prefix, followed by the time information. Such filenames # would be too large for PCs running DOS, therefore there is no prefix # for the USEARCSSCONCEPT (write NA in the appropriate column). # # Often times you will set the time lags to UK and run everything through # converall determining the actual time lag. From this output you can # make a statistics and determine an average time lag which you than can # place in this table and reprocess the whole data after recompiling # convertall with FINDACTUALLAG set to FALSE in convertall.h # Normally the actual lag should be the best value to use, but for # comparisions with people who only can process their data with a fixed # time lag, this could be useful (or if the time lag seems to freak out # when you run convertall with FINDACTUALLAG set TRUE). # BEGINNING ENDING TZ Z H0 D0 DIR AN LAGS....... PREFIX LOCATION 01.01.2000 00:00:00 31.05.2000 23:59:59 MEZ 7.00 0.3 0.2 0 0 NA Kerzersmoos FM-Test 31.05.2000 00:00:00 31.12.2000 23:59:59 MEZ 31.50 19.0 12.5 0 0 NA BAYREUTH-Finimsas 01.01.2000 00:00:00 31.12.2000 23:59:59 MEZ 31.50 19.0 12.5 0 0 d Fichtel FM-Thomas 01.05.2002 00:00:00 31.12.2004 23:59:59 MEZ 31.50 19.0 12.5 0 0 NA CARBOMONT Seebodenalp 01.01.1995 00:00:00 09.08.2005 09:30:00 MEZ 36.0 25.0 20.0 235 2 UK UK Davos Davos Seehornwald WSL 10.08.2005 12:31:00 19.12.2006 23:59:59 MEZ 36.0 25.0 20.0 239 0 NA Davos Seehornwald ETH R2A 20.12.2006 00:00:00 31.12.2010 23:59:59 MEZ 36.0 25.0 20.0 319 0 NA Davos Seehornwald ETH R3-50 # UK = time lag unknown (but analog channel was used) # NU = data channel was not used # NA = not available, data not provided or unknown value