September 25, 1997
Subject: the X-Roots workshop discussion on climate datasets.
To: Caroline & David
Before the standard variable names become too standard, we should look at more
possibilities and ensure that our format is flexible. Given the example
MATMPM as a pattern for discussion, could we define that the first letter will
be the time scale for averaging/sampling/totalizing? Thus:
A annual,
S season, or possibly G for Growing and N for Dormant or Non-growing,
M monthly,
D daily,
H Hourly.
Because some variables are averaged and some are totaled, the final letter
could be:
M mean,
T total,
S sample, or O for observation which is bad because of potential O-0
confusion.
Maximum and minimum should not be in the final position because we could have
the monthly mean of the daily maximums OR the absolute maximum for the month.
For that latter case:
E extreme value for the time period.
This leaves the center of the parameter name to designate the important thing,
WHAT IS IT? I presume that we want a standard character length for these
names and you have chosen a total of 6. However, the suggested name for
precipitation violates my presumption. Ignoring that, and keeping in mind the
max-min comment above:
ATAV air temperature,
ATMN air temperature minimum,
ATMX air temperature maximum,
STxx soil temperature where xx=depth in cm,
PREC all precipitation,
PRSD snow depth,
WDSP wind speed,
WDAZ wind direction,
WVEC wind vector,
GSOL global solar,
DSOL direct beam solar,
GPAR photosynthetically active radiation,
REHU relative humidity,
AVPR actual vapor pressure,
VPDE vapor pressure deficit,
DEWP dew point,
EVAP pan evaporation,
and probably other variables. The point is that we could handle a lot of
different Level 3 data with such a system.
This last exercise raises a concern that the V-Many table structure could get
very large for some sites or studies, and, authoritarian dictatorship would be
necessary if someone wanted all the table columns to be uniform and in the
same order across all sites. Because Level 3 data tends to be study-oriented
and not always long-term in the continuing sense, V-One seems the best
structure for these data.
Lloyd
--------------------+-------------------------------+---------------------
Lloyd W. Swift, Jr. | Coweeta Hydrologic Laboratory | 3160 Coweeta Lab Rd
704-524-2128 | USDA, FS, | Otto NC 28763
fax: 704-369-6768 | Southern Research Station |
internet: | /S=L.SWIFT/OU1=S29L05A@mhs-fswa.attmail.com
| lswift@sparc.ecology.uga.edu |
| LSwift@LTERnet.edu |
| lwswift@dnet.net |
--------------------+-------------------------------+---------------------