Continuous:Hydrometeorological Data

From Gsshawiki
Jump to: navigation, search

In order to perform continuous simulation in GSSHA hydrometeorological data are required for the entire period of the continuous simulation. The required data are hourly values of:

  • barometric pressure,
  • relative humidity,
  • total sky cover,
  • wind speed,
  • dry bulb temperature,
  • direct radiation, and
  • global radiation.

These data are available from a wide variety of sources and three different input file formats can be used. The GSSHA model also contains provisions to synthetically produce solar radiation estimates and fill in short term data gaps.

9.3.1 Hydrometeorological Input File Formats

The three file formats that GSSHA supports are discussed below in order of preference. Simple file formats are preferred because the likelihood of a data format error is reduced. The WES format is the simplest because it uses no alphabetical characters. Users are encouraged to produce and use hydrometeorological data files in the WES format for continuous simulations. The other formats contain data quality flags, extraneous data, and occasionally errors. Such errors can be very difficult to locate. Reducing data down to the WES format through independent software (such as a spreadsheet) ensures that the data are properly formatted and contains no extraneous values. HMET_WES (Recommended format)

This file contains only hourly values of the required hydrometeorological parameters: barometric pressure, relative humidity, total sky cover, wind speed, temperature, direct radiation, and global radiation. The HMET_WES format contains the following numbers in columns 1-11 (Table 15):

Col. Variable Units No Data Flags Type
1 Year (4 digit)     integer
2 Month     integer
3 Day     integer
4 Hour     integer
5 Barometric Pressure in Hg No Data (ND) = 99.999 real
6 Relative Humidity  % ND=999 integer
7 Total Sky Cover  % ND=999 integer
8 Wind Speed kts ND=999 integer
9 Dry Bulb Temperature °F ND=999 integer
10 Direct Radiation W h m-2 ND=9999.99 real
11 Global Radiation W h m-2 ND=9999.99 real

Table 15 – HMET_WES file format

Data format is important; the year is stored as a 4 digit number; the hour varies between 0 and 23. The data in columns 1-4, 6-9 are stored as integers, while the data in columns 5, 10, and 11 are stored as real numbers. The data in this file should be space delimited. The number of spaces between columns does not matter, provided that the total width of the line is less than 256 characters. There must be 1 hourly record for each hour of the intended duration of the continuous simulation, even if there are no data available for that hour. Gaps in the hourly record will cause the GSSHA simulation to terminate prematurely.

The following are three example lines from a HMET_WES formatted hydrometeorological data file (note the distinct presence of both real and integer numbers):

1995    4   14   23   29.625    86   100   000    53   9999.99   9999.99
1995    4   15    0   29.625    83   100   003    53   9999.99   9999.99
1995    4   15    1   29.615    90    80   003    52   9999.99    9999.99

This data input format is strongly recommended because it is compact, easy to read and write and contains no extraneous information. The no data flags are important. GSSHA has a set procedure for dealing with missing data. HMET_SAMSON (Recommended over surface airways format)

In 1993 the National Climatic Data Center (NCDC) published a CD-ROM containing surface meteorological observations from 1961-1990. This CD-ROM is the result of collaboration between the National Oceanic and Atmospheric Administration (NOAA) and the U.S. Department of Energy, National Renewable Energy Laboratory. The SAMSON CD-ROM represents an excellent nationwide dataset for the 1961-1990 period. Unfortunately, this CD-ROM is not being updated as new data become available.

The following is an example of data from the SAMSON CD-ROM. A reader program provided with the SAMSON data produces this data format. The ASCII Samson data format includes a one line station header (in this case the data are from the Memphis, Tennessee station), and a one line descriptive header with column ID’s. These two header lines are followed by a user selectable number of hourly data points (in this example two hourly lines are shown).

~13893 MEMPHIS                TN  -6  N35 03  W089 59    87
~YR MO DA HR I    1    2       3       4        5       6  7       8     9   10     11   12    13     14       15     16   17         18   19        20  21
   82   4      1      1 0    0    0    0 ?0    0 ?0    0 ?0  0  0  12.8   6.1  64 1011 100  2.1   24.1  77777 999999999   16 99999.    0  26
   82   4      1      2 0    0    0    0 ?0    0 ?0    0 ?0  0  0  12.2   6.1  67 1011   80  2.1   24.1  77777 999999999   16 99999.    0  26

The first header line consist of the Weather Bureau Army Navy (WBAN) number for the station (13893), followed by 26 character descriptive string (MEMPHIS), the state name where the station is located (TN), the time difference between local standard and UTC in hours (-6), degrees and minutes of latitude, N|S, (N35 03), degrees and minutes of longitude ,E|W, (W089 59) and the station elevation in meters above sea level (87). The first header record containing the station information was written by NCDC with the following FORTRAN format statement:


The second line is for the identifier record, which shows the numerical identifier for the data elements in the National Solar Radiation Data Base (NSRDB) Synoptic Format (see:

The observed data occupy the third record and all ensuing records (lines) in the file until a new year of data occurs. When a new year of data is encountered the header record and identifier record are repeated, followed by more data records. The data records include year, month, day, hour (1-24) in local standard time. The number below the “I” column is an “observation indicator” for which 0 means the weather observation was made, and a 9 indicates that the observation is missing. The following table describes the contents of each data record (Table 16):

No. Name Value Range Description and Notes:
YR Year (61-90) Year of observation
MO Month (1-12) Month of observation
DA Day (1-31) Day of month
HR Hour (1-24) Hour of the day, local standard time.
I Observation Indicator 0
0= weather observation made
9= no observation made
NOTE: if this field=9 OR if field 13 (wind speed) = missing (9999. or 99.0) then fields 6,7,8,10,11,17, and 18 were all modeled and not actually observed.
1 Extraterrestrial Horizontal Radiation 0-1415 Amount of solar radiation in W-h m-2 on a horizontal surface at the top of the atmosphere during the 60 minutes preceding the hour indicated.
2 Extraterrestrial Direct Normal Radiation 0-1415 Amount of solar radiation in W-h m-2 received on a surface normal to the sun at the top of the atmosphere during the 60 minutes preceding the hour indicated.
3 Global Horizontal Radiation 0-1415
flag for data 0-9
9999=missing data
Total amount of direct and diffuse solar radiation in W-h m-2 on a horizontal surface during the 60 minutes preceding the hour indicated.
4 Direct Normal Radiation 0-1415
flag for data 0-9
9999=missing data
Amount of solar radiation in W-h m-2 received within 5.7 degrees field of view centered on the sun during the 60 minutes preceding the hour indicated.
5 Diffuse Horizontal Radiation 0-1415
flag for data 0-9
9999=missing data
Amount of solar radiation in W-h m-2 received from the sky (excluding the solar disk) on a horizontal surface, during the 60 minutes preceding the hour indicated.
6 Total Sky Cover 1-10
99=missing data
Amount of sky dome (in tenths) covered by clouds
7 Opaque Sky Cover 1-10
99=missing data
Amount of sky dome (in tenths) covered by clouds that prevent observing the sky or higher cloud layers.
8 Dry Bulb Temperature -70.0 to +60.0
9999.=missing data
Dry bulb air temperature in degrees C.
9 Dew point -70.0 to +60.0
9999.=missing data
Dew point temperature in degrees C.
10 Relative Humidity 0-100
999=missing data
Relative humidity in percent.
11 Station pressure 700-1100
9999=missing data
Station barometric pressure in mb.
12 Wind Direction 0-360
999=missing data
Wind direction in degrees. N=0 or 360, E=90, S=180, W=270
13 Wind Speed 0.0-99.0
9999. or 99.0= missing data
Wind speed in m s-1
14 Visibility 0.0-160.9
99999.=missing data
Horizontal visibility in kilometers.
15 Ceiling height 0-30450
999999=missing data
Ceiling height in meters
16 Present weather Table denoted by 9 indicators. Present weather conditions. (See NOAA document for codes).
17 Precipitable water 0-100
9999=missing data.
Precipitable water in millimeters
18 Broadband aerosol optical depth 0.0-0.900
99999.=missing data
Broadband aerosol optical depth (broadband turbidity) on the day indicated.
19 Snow depth 1-100
9999=missing data
Snow depth in centimeters on the day indicated.
20 Days since last snowfall 0-88
88=88 or > days
999=missing data.
Number of days since last snowfall
21 Hourly precipitation 000000-099999 In inches and hundredths.

Table 16 – HMET_SAMPSON file format

The SAMSON CD-ROM is available by contacting the NCDC at:
Climatic Services Branch
National Climatic Data Center
Federal Building
Asheville, NC 28801-2696
Phone - (704) 271-4800
Fax - (704) 271-4876 HMET_SUFAWAYS - NOAA/NCDC Surface Airways Format (Not recommended)

The Surface Airways data format is included in the list of supported formats because surface meteorological data ordered from NCDC are delivered in this format. In the Surface Airways data format each data element (wind, temperature, radiation, etc.) is written on a separate line in the file. Each line contains 24 hourly values. Missing data are not flagged.

The details of this data format are discussed in “TD-3280, Hourly Surface Airways Observations”, published by the National Oceanic and Atmospheric Administration, National Climatic Data Center. GSSHA users are strongly encouraged to reformat Surface Airways observations into the WES data format. There is a strong likelihood that data in the Surface Airways format will produce problems during the GSSHA simulation. This occurs not as consequence of faulty data, but because of the large number of data flags in the Surface Airways output format make the Surface Airways data format difficult to read in a robust fashion.

9.2.2 Missing Hydrometeorological Data

Frequently, hydrometeorological records have periods of missing data because observational instruments are inoperative, the data are not reliable, or certain observations (e.g. direct radiation, global radiation) are not available for a given station. Because hourly hydrometeorological inputs are required for the entire simulation period, missing data are synthetically produced within the GSSHA model. Radiation Data

Many hydrometeorological stations do not record solar radiation measurements. Missing solar radiation data are simulated in GSSHA based on the location of the watershed, day of the year, and time of day, as described in Section 9.1. When only a portion of the solar radiation data is missing, it is replaced as described in the next section. Short Duration Gaps in Hydrometeorological Measurements

When required data are missing, GSSHA creates the needed data by either persistence or persistence adjusted for the time of day. Missing variables without strong diurnal variability are filled in by “persistence” estimates. These variables include:

  • barometric pressure,
  • wind speed,
  • dew point temperature and,
  • percent cloud cover.

Persistence means that the value is held constant until a new observation becomes available. Missing variables with strong diurnal components:

  • air temperature,
  • global radiation,
  • and direct radiation.

are replaced with the last good reading from the same time of day. It is the responsibility of the user to check for missing data. Extended periods of missing record (many days) will cause GSSHA to simulate conditions for the last day of good record over and over again. This scheme requires that the hydrometeorological data file begin with at least one day of good observations of all required variables.

GSSHA User's Manual

9 Continuous
9.1     Computation of Evaporation and Evapo-transpiration
9.2     Computation of Soil Moisture
9.3     Hydrometeorological Data
9.4     Snowfall Accumulation and Melting
9.5     Sequence of Events During Long-Term Simulations