National Climatic Data Center
DATA DOCUMENTATION
FOR
DATA SET TD-6421
Enhanced hourly wind station data for the contiguous United States
Version 1.1
December 6, 2001
Pavel Ya. Groisman
National Climatic Data Center
151 Patton Ave.
Asheville, NC 28801-5001 USA
Table of Contents
TOPIC PAGE
________________________________________________________________
INTRODUCTION
1. Data Set ID...................................... 3
2. Data Set Name.................................... 3
3. Data Set Aliases................................. 3
________________________________________________________________
DESCRIPTION
4. Access Method and Sort for Archived Data......... 3
5. Access Method and Sort for Supplied Data.........11
6. Element Names and Definitions....................11
7. Start Date.......................................12
8. Stop Date........................................12
9. Coverage........................................ 12
10. Location........................................ 12
11. Keyword......................................... 12
12. How to Order Data............................... 12
________________________________________________________________
DATA CENTER
13. Archiving Data Center........................... 12
14. Technical Contact............................... 12
________________________________________________________________
DATA QUALITY
15. Known Uncorrected Problems...................... 13
16. Quality Statement............................... 14
17. Revision Date .................................. 14
_______________________________________________________________
OTHER DATA SETS
18. Source Data Sets................................ 14
19. Essential Companion Data Sets................... 14
20. Derived Data Sets............................... 15
________________________________________________________________
SUMMARIZATION
21. References...................................... 15
22. Summary......................................... 15
1. Data Set ID: 6421
2. Data Set Name: Enhanced hourly wind station data for the contiguous United States
3. Data Set Aliases: None
4. Access Method and Sort for Archived Data:
Data are stored as several groups of ASCII files :
4.1. First group of files contains station metadata:
"US_1657stns.lst", “US_1657stns.lst_extended", “metadata_NCDC_stn.lst”, “exist-metadata_files”, and “anem_elev_inf”.
The original list of stations with hourly information is collected in files "US_1657stns.lst” and "US_1657stns.lst_extended". The format of these two files is similar except:
File "US_1657stns.lst" contains supplementary text information in lines after 1662 and in positions after 70 that identify composite stations and provides some additional information how these stations were handled. First four lines in "US_1657stns.lst" have clarification information about the column meanings and last lines (after line 1662) have comments related to different merging steps that were made to few dozen of station files during their pre-processing. File "US_1657stns.lst_extended" does not contain this information and have strictly 1657 lines with information (line per station). It (a) has an additional bit of information in the end of each line (blank, or '1') that indicate if the station was in TD3280 at least for a part of the period ('1' stays for "yes" in this inquiry) (b) shows a first and last years with the data for a given station, (c) shows a regional partition of the station within the contiguous U.S. (1 through 9) used in Groisman et al. (2001), and (d) has information about the date of ASOS installation at the station (if any).
These two files can be accessed with a FORTRAN program using the following code fragment:
real lat,lon
integer elev,yearb,yeare,nreg,tz
character*1 li, first
character*6 wmo6,coop
character*5 wban
character*19 name
character*13 linea
11 format(a1,1x,a6,1x,a19,1x,a2,1x,f7.2,f6.2,1x,i4,
*1x,a5,1x,a6,1x,i3,1x,a1,2i5)
for file "US_1657stns.lst" (lines 5 through 1662)
read(16,11)li,wmo6,name,st,lon,lat,elev,wban,coop,tz,
* first,yearb,yeare
for file "US_1657stns.lst_extended" :
read(16,11)li,wmo6,name,st,lon,lat,elev,wban,coop,tz,
* first,yearb,yeare, nreg, linea
Variables used in the code fragment are as follows:
li is an index (with values ‘1 ‘or ‘2’) that characterizes the number of stations that historically had this ID;
wmo6 is an extended (6-digit) "wmo6" number used by US air force for station identification;
name is the station name (up to 19 characters);
st is a two character state abbreviation;
lon and lat are longitude and latitude (negative values mean west and south);
elev is the station elevation in meters;
wban is 5-digit U.S. National Weather Bureau number (WBAN number if present; otherwise blank);
coop is a 6-digit U.S. cooperative station number of the station (if present, otherwise blank);
tz is a time zone indicator relative to UTC (so EST is -5 UTC and PST is -8);
first is a one-character indicator of the first order station (set to ‘1’, otherwise blank; absent in “US_1657stns.lst”);
yearb and yeare are the first and the last year with valid data in the archive (absent in "US_1657stns.lst");
nreg is a region number in the following order: Northwest, Missouri River Basin, Upper Mississippi, Northeast, California and Nevada, Southwest, South, Midwest, and Southeast (variable is absent in “US_1657stns.lst”; map of this partition is shown in Groisman et al. [2001]);
linea string is blank if ASOS was not installed at the station (as of July 31, 2001); otherwise first four characters of this string are set to ‘ASOS’ and the last 8 characters contain the date (year, month, day) of the ASOS comissioning that can be read in format (i4,2i2.2); this string is absent in “US_1657stns.lst”.
Example of file "US_1657stns.lst_extended”:
….
1 722025 FT LAUDERDALE/HOLLY FL -80.15 26.07 7 12849 -5 1972 2000 9 ASOS 19980909
1 722026 HOMESTEAD AFB FL -80.38 25.48 2 12826 -5 1943 2000 9
1 722027 MIAMI BEACH (CGS) FL -80.13 25.77 1 -5 1983 1984
……
The list of stations with adjusted wind information is provided in “exist_metadata_files” that is a subset (1304 lines) of the "US_1657stns.lst_extended”.
File " metadata_NCDC_stn.lst" contains information about the stations with complete metadata digitized at NCDC (1220 lines)
and can be accessed with a FORTRAN program using the following code fragment:
character*4 csign,ASOS
character*7 lat1
character*8 lon1
character*5 elev1,wmo, wban1,wban2
character*35 name1
integer yeara,mona,daya
read(11,3)wmo6,csign,lat1,lon1,elev1,name1,
* wmo,ASOS,yeara,mona,daya,wban1,wban2
3 format(a6,1x,a4,1x,a7,1x,a8,1x,a5,1x,a35,1x,a5,
* 1x,a4,1x,i4,2i2.2),2(1x,a5))
Variables used in the code fragment are as follows:
(a) wmo6 is the extended (six character) WMO station number ;
(b) csign is the station call letters (if exists, otherwise blank);
(c) lat,lon are the station coordinates (lat. and long.);
(d) elev1 is the station elevation (in feet !!!);
(e) name1 is the station name (up to 35 characters);
(f) wmo is the true WMO # (if exists; otherwise blank);
(g) ASOS is a 4-character indication if the station became ASOS (or AWOS1) type; and
(h) yeara, mona, and daya compose the date of the ASOS comissioning;
(i) wban1 and wban2 up to two WBAN numbers assigned to the station (if any).
Example:
.....
722313 LIX N 30 20 W 89 49 27 SLIDELL, LA 72233 12916
722314 ARA N 30 02 W 91 53 24 NEW IBERIA, LA ASOs 19980505 03934
File “anem_elev_inf” (1305 lines) was generated from several supplementary files (available upon request). File "elev_NCDC_stns" contains information about the dates of the anemometer elevation changes for all stations with complete metadata digitized at NCDC. The same structure of the anemometer elevation information files was generated for the 260 U.S. first order stations of contiguous U.S. and Alaska [ file "elev_1st_order_stns"] and for U.S. Air Force and Navy stations metadata files "elev_air_force_stns" and "elev_navy_stns". Finally, all this elevation information was combined into the file "merge_anem_elev_inf.OUT" in the order of preference (trustworthiness): First are "elev_1st_order_stns", then are "elev_NCDC_stns", then "elev_navy_stns" and "elev_air_force_stns". A separate file with duplicate information "metadata_files_duplicates" was generated and was used to verify the metadata information about the anemometer elevation that comes from different sources. The discrepancies in this information were analyzed and corrected.
Final file “anem_elev_inf” combines all these files without duplicates and can be accessed with a FORTRAN program using the following code fragment:
character*6 wmo6
integer i, nmbr, year(30),mon(30),iday(30)
real elevf(30)
read(12,6)wmo6,nmbr,(year(i),mon(i),iday(i),elevf(i),i=1,nmbr)
6 format(a6,1x,i2,30(1x,i4,2i2.2,f5.1))
Variables used in the code fragment are as follows:
For each station a line of information contains:
(a) the extended (six-character) wmo6 station number ;
(b) the number of changes at the station, NMBR;
(c) up to 30 groups of changes:
Date of change: Year, month, day (year(i),mon(i),iday(i)) and
A new anemometer elevation in feet (elevf(i)).
If the anemometer elevation is not known then it is indicated to be 999.0. If it is known that the AWOS installation was implemented but the anemometer elevation was not reported, then it was assumed to be 10 m (or 33 ft). Complete information about the anemometer elevation at all U.S. ASOS sites was kindly provided by Richard Parry (NOAA/NESDIS). When unknown elevation or day of change were met within the station history record when valid wind speed data exist, following assumptions were used: (1) If the day of change is not known, it is set to the 15th of the month. (2) During the processing of the metadata file for each station, we assumed that the anemometer elevation prior to the first and after the last record of its changes remain the same as indicated in these first and last records respectively. (3) When adjusting wind speed values for the period with unknown anemometer elevation (elevf(i) = 999.0), elevf was assumed to be typical for the year when the change to this unknown elevation was documented: 50 feet (year<1951), 20 feet (1959< year <1995), and 30 feet otherwise.
Example of file “anem_elev_inf”:
.....
723125 4 19420915999.0 19510119 30.0 19671115 20.0 19981104 33.0
723126 3 19430715999.0 19440807 39.0 19520415 49.0
723139 8 19460415999.0 19461212 47.0 19550410 56.0 19570301 57.0 19570715 33.0 19601115 32.0 19790415 33.0 19950701 32.8
723140 6 19340810 86.0 19510426 41.0 19540804 58.0 19601117 20.0 19821116 33.0 19980701 32.8
723143 3 19620924 40.0 19720402 35.0 19900111 33.0
723145 7 19480203 30.0 19490528 32.0 19551215 33.0 19700806 22.0 19860411 33.0 19970827 26.0 19970905 32.8
723150 3 19310101 92.0 19640901 20.0 19960601 26.0
.....
All files in this group are sorted in ascending order by wmo6 key.
4.2. Second group of files contains results of quality checking
When we saw that the extreme wind speed value is unbelievably high compared to those during the neighboring hours, at the neighboring sites, and under the weather conditions in the area (atmospheric pressure and temperature gradients, cloudiness, weather codes, and presence of hurricane/tropical storm/depression nearby were considered), we flagged this value as "questionable" (quality flag "Q"). Quite often during this analysis, we found that the value most probably has a wrong scale (i.e., multiplied by 10) and scaling back will return it to a proper value similar to those in the neighborhood. We flagged these values as "with scale-error" (flag "S”).
We have generated a list of the flagged wind speed values partitioned into two files "wind_errors" and “wind_extremes” that were created and can be accessed with a FORTRAN program using the following code fragment:
real lat, lon, extre
integer elev
character*1 flg1,flg2,flg3
character*6 wmo6
character*2 st
character*19 name
character*10 lin
read(19,223)wmo6,name,st,lon,lat,elev,
*lin,extre,flg1,flg2,flg3
223 format(a6,1x,a19,1x,a2,1x,f7.2,f6.2,1x,i4,1x,
*a10,1x,f5.1,1x,3a1)
where the station ID (wmo6), name, state (st), coordinates (lon, lat) and elevation (elev, in meters) are accompanied with the date of extreme (year, month, day, hour coded in the string of characters lin in internal format (i4,3i2.2)) and its value, extre, in m sec-1. Flg3 is an aforementioned quality flag and flg1 and flg2 indicate the source of the data when both, TD3280 and DATSAV3, data streams were available. In the example below, *1 for station 722280 BIRMINGHAM indicates that the erroneous data was in DATSAV3 data stream and was skipped in TD3280.
Example:
722275 ANDALUSIA/OPP ARPT AL -86.40 31.32 94 2000090609 73.5 S
722268 DOTHAN MUNICIPAL AL -85.45 31.32 122 2000122821 62.2 S
722268 DOTHAN MUNICIPAL AL -85.45 31.32 122 1996122414 56.0 S
722280 BIRMINGHAM MUNI AL -86.75 33.57 192 1973050805 56.0 *1Q
722275 ANDALUSIA/OPP ARPT AL -86.40 31.32 94 2000011415 55.0 S
.....
A slightly expanded version that combines information of these two files: "wind_problems.1st_run" contains an additional line for each error/extreme record that may provide some justification for the flg3 selection and period of record used for analysis and the values that were finally selected as valid extremes (same as in file “valid_extremes”). Each of these valid extremes have explanations and are flagged as "justified extreme value" (flg3 = "E") and/or "valid individual extreme value that does not have a sufficient level of support by neighboring sites and/or hours" (flg3 ='U"). Example of the part of this file is below:
.....
NY 744865 30 4 1943 2000
744865 WESTHAMPTON BEACH NY -72.63 40.85 20 1983011817 32.4 Q
NY 725027 30 6 1975 1996
725027 EAST MORICHES (CGS) NY -72.75 40.78 1 1981062707 31.8 S
NY 725030 30 7 1948 2000 SOME EVIDENCE OF THE WIND GUST?
725030 NEW YORK/LA GUARDIA NY -73.90 40.77 9 1979070508 31.8 *1Q
NY 744865 30 3 1943 2000 HURR. CAROL; CENTER 37.3 -74.2; SPEED 85 KNOTS
744865 WESTHAMPTON BEACH NY -72.63 40.85 20 1954083108 32.4 E
NY 724098 30 5 1975 1987 SUPPORTED BY NEIGHBOR HOURS AND SITES
724098 AMBROSE/FT TILDEN NY -73.82 40.45 1 1984032907 31.8 E
NY 994100 30 8 1985 1997 SUPPORTED BY NEIGHBOR HOURS AND SITES
994100 AMBROSE (LGT-SHIP)& NY -73.83 40.47 0 1992121113 31.8 E
NY 994100 30 9 1985 1997 SUPPORTED BY NEIGHBOR HOURS AND SITES
994100 AMBROSE (LGT-SHIP)& NY -73.83 40.47 0 1993031316 31.8 E
NY 725038 30 10 1942 2000 SUPPORTED BY NEIGHBOR HOURS AND SITES
725038 NEWBURGH/STEWART NY -74.10 41.50 150 1950112518 31.3 E
.......
Here some extreme values in New York state are considered valid (flag E) and are explained why.
An additional group of output files: "wind_extremes.2d_run.TRUSTED" and "wind_extremes_extended.2d_run.TRUSTED" can be useful for analyses of extremes and accompanied weather conditions. The first file contains valid 10 extreme values for each state those already identified in "wind_problems.1st_run".
File "wind_extremes_extended.2d_run.TRUSTED" provides for each of this extreme hourly wind record the entire daily suite of hourly values in .file2 format for the station with extreme values and for up to six nearest locations that had observations in a given year. The “.file2” format is an internal format of the original data that include atmospheric pressure, temperature, humidity, cloudiness, and weather information. Additional output files (available by special request due to large volume of supplementary data included in these files) provide information similar to that available in "wind_extremes_extended.2d_run.TRUSTED" for the majority of rejected extreme values from "wind_problems.1st_run". Exceptions are only for those rejected values that were found by pattern during an extensive search within the station with questionable extreme values. In these cases, false "extreme" was flagged but no neighborhood weather information for it is provided.
The data with *.file2 format can be accessed with a FORTRAN program using the following code fragment:
integer month,year,day,hour,wea(10),cldtyp(10),
* nwx,ncld,maxof,wndir,locc,opac
real pres,temp,vp,rh,wndspd
read(10,2)
* year,month,day,hour,pres,temp,vp,rh,
* maxof,wndspd,opaq,tcc,wx,c37,wndir,locc,ncld,nwx,
* (cldtyp(i),i=1,ncld),(wea(j),j=1,nwx)
2 format(i4,3i2,f5.1,2f4.1,f3.0,i2,f4.1,3i2,a1,4i3,30i2)
where: year, month, day, and hour, are followed by
atmospheric pressure at the station (pres in hPa), surface air temperature (temp in ?C),
water vapor pressure (vp in hPa), relative humidity (rh in percent),
maximum of cloudiness and weather codes (maxof),
wind speed (wndspd, in meters per second),
opacity (opac) and total cloud cover (tcc) in hundredths,
weather code at observation time (wx), description of these codes is available in TD 3280,
code indicating which cloud element was used for cloud information, c372,
wind direction (wndir) in degrees from N clockwise,
low cloud cover (locc) in hundredths,
number of cloud types to follow (ncld), number of weather codes to follow (nwx), and finally,
variable number of two digit numbers - cloud types (cldtyp) followed by weather codes (wea). Use documentation for US. Surface Airways Hourly, TD 3280, for cloud type and weather types codes definitions.
It should be noted that the data in file " wind_extremes_extended.2d_run.TRUSTED " being compiled with .file2 format are separated by string with station name, WMO, coordinates and qualifier that indicates if the station is a member of checking of suspicious “extreme” value or a member of the neighbor group. Qualifier “ neighbor” is explicitly printed at the end of this string.
Example:
First two lines of file wind_extremes.2d_run.TRUSTED
AL 722269 1 1 1954 2000
722269 CAIRNS AAF/OZARK AL -85.72 31.28 91 1995100418 33.4
.......
Corresponding information from file “wind_extremes_extended.2d_run.TRUSTED”
in .file2 format. Station EGLIN AFB supports extreme values at CAIRNS AAF.
AL
722269 CAIRNS AAF/OZARK AL -85.72 31.28 91 1995100418 33.4
199510040099999 233 277 9699 46-199 01280 -1 1 11810
199510040199999 227 27710099 25-199 01120 -1 0 110
199510040299999 227 27710087 25-187 01100 -1 0 110
199510040399999 227 27710099 20-199 01 90 -1 1 11525
199510040499999 227 27710099 36-199 01 80 -1 0 125
199510040599999 233 28610099 30-199 01 60 -1 0 125
199510040699999 233 28610099 36-199 01 70 -1 1 11525
199510040799999 238 29610099 46-199 01 80 -1 0 125
199510040899999 238 29610099 41-199 01 80 -1 0 125
199510040999999 238 29610099 51-199 01 70 -1 2 1182225
199510041099999 244 296 9699 61-199 01 80 -1 0 125
199510041199999 250 306 9699 61-199 01 80 -1 0 125
199510041299999 250 306 9699 72-199 01 80 -1 1 11825
199510041399999 250 306 9699 77-199 01100 -1 0 110
199510041499999 250 306 9699 72-199 01 90 -1 0 120
199510041599999 261 317 9499 87-199231 90 -1 1 2151023
199510041699999 255 306 9499 97-199 01100 -1 0 110
199510041799999 255 306 9499 108-199 01100 -1 0 110
199510041899999 255 306 9499 334-199 01160 -1 1 11810
199510041999999 255 306 9499 257-199 01160 -1 0 110
199510042099999 250 296 9399 128-199231160 -1 0 21023
199510042199999 250 296 9399 128-199 01160 -1 1 11523
199510042299999 244 286 9499 128-199 01200 -1 0 123
199510042399999 244 286 9499 128-199 01200 -1 0 123
.............
722210 EGLIN AFB/VALPARAIS FL -86.53 30.48 26 neighbor
199510040099999 227 27710099 51-199 01 40 -1 2 216222551
199510040199999 227 27710099 56-199 01 20 -1 0 22551
199510040299999 227 27710099 51-199 01 40 -1 0 22551
199510040399999 227 27710099 46-199 01 50 -1 1 2162551
199510040499999 233 28610099 51-199 01360 -1 0 21086
199510040599999 238 29610099 56-199 01 50 -1 0 22551
199510040699999 244 30610099 41-199 01 60 -1 0 22551
199510040799999 250 31710099 66-199 01 60 -1 0 22551
199510040899999 250 31710099 82-199 01 60 -1 0 22551
199510040999999 255 317 9799 87-199 01 60 -1 1 2132551
199510041099999 255 317 9799 102-199 01 90 -1 0 22551
199510041199999 261 327 9799 77-199 01 90 -1 0 22551
199510041299999 261 327 9799 128-199 01 90 -1 1 2132551
199510041399999 266 327 9499 123-199 01 90 -1 0 22551
199510041499999 255 32710099 149-199 01100 -1 0 22551
199510041599999 244 30610099 226-199 01 90 -1 1 11325
199510041699999 244 306100 0 159-1-1 0 110 -1 0 125
199510041799999 250 306 96 0 360-1-1 0 110 -1 0 125
199510041899999 250 306 9699 205-199 01110 -1 0 22551
199510041999999 250 268 8599 205-199 01200 -1 0 22570
199510042099999 238 29610099 205-199 01190 -1 0 125
199510042199999 238 286 9799 144-199 01270 -1 1 2132551
199510042299999 233 28610099 128-199 01240 -1 0 22551
199510042399999 233 242 8499 87-199 01260 -1 0 22551
Two additional quality flags were inserted into original wind data using the hurricane and tropical storm statistics. We flagged each wind speed value that is originated from the site that is 300 km away from the hurricane center [ flag ‘H’] or 150 km away from the tropical cyclone center [flag ‘T’].
4.3. Hourly wind data with quality flags inserted after the QC process; unadjusted (*.wnd1) and adjusted (*.wnd2) extensions
The asterisks in the file names indicate a 6-digit extended WMO code of the station (wmo6).
There are 1657 individual files with unadjusted wind hourly information (two of them, 745203.wnd1 and 744920.wnd1 contain no valid wind information at all) but only 1304 files with adjusted wind information. The list of stations with adjusted wind information is provided in ““exist_metadata_files” file described in 4.1.
Each individual file can be accessed with a FORTRAN program using the following code fragment:
integer Sp,Dr, Sp10, Sp_surf, year,month,day,hour
character *1 S,Q,Z
33 format(a6,i4,3i2.2,i4,1x,i3,2a1,i4,1x,f4.1,a1,i4)
34 format(a6,i4,3i2.2,i3,1x,i3,2a1)
For file with extension (*.wnd1):
read(16,34) wmo6,year,month,day,hour, Sp,Dr,S,Q
For file with extension (*.wnd2):
read(16,33)wmo6,year,month,day,hour,Sp,Dr,S,Q,Sp10,An,Z,Sp_surf
Variables used in the code fragment are as follows:
Station ID (wmo6) is followed by:
* date (year, month, day, hour in local standard time, LST),
* wind speed at the anemometer height, Sp (in m sec-1 x 10),
* wind direction, Dr (degrees),
* flag of the data source, S (it is equal to ‘1’, i.e., originated from the NCDC TD 3280 archive, and blank when the data were originated from DATSAV3),
* quality flag, Q3,
* adjusted to 10 m above the ground wind speed, Sp10 (in m sec-1 x 10),
* anemometer height, An (m),
* flag of the presence of snow on the ground, Z (blank when snow is absent, ‘1’ when snow on the ground is present but less than 10 cm, and ‘2’ when snow on the ground is above 10 cm; these flags define different surface roughness parameter, Z0, in the wind adjustment formulae), and
* wind speed adjusted to 10 m above the surface, Sp_surf (in m sec-1 x 10 that is provided only when snow on the ground is above 10 cm, e.g., for Sault Ste Marie in the example below).
Missing values for wind speed and direction are coded with negative values -0.5 and –1 respectively.
Example of the hourly wind information for Miami (file = 722020.wnd2) and Sault Ste Marie (file = 727340.wnd2). The file fragments for these stations with extension “.wnd1” for these stations are the same but do not have Sp10, An, Z, and Sp_surf columns.
A. Miami, Florida
wmo6 Yr M D H Sp Dr SQ Sp10 An Z Sp_surf
.........
7220201948100416 15 1351 14 18.9
7220201948100417 20 901 18 18.9
7220201948100418 15 451 14 18.9
……
7220201948100515 118 681H 106 18.9
7220201948100516 77 1131H 69 18.9
7220201948100517 216 1131H 195 18.9
7220201948100518 303 1131E 273 18.9
7220201948100519 15 3601H 14 18.9
7220201948100520 221 3151H 199 18.9
7220201948100521 180 3151H 162 18.9
………..
B. Sault Ste Marie, Michigan
7273401968122122 30 1001 32 6.11
7273401968122123 -5 -11 -5 6.11
7273401968122200 -5 -11 -5 6.12 -5
7273401968122201 36 901 39 6.12 39
7273401968122202 -5 -11 -5 6.12 -5
…………
7273401968122216 72 1101 77 6.12 77
7273401968122217 -5 -11 -5 6.12 -5
7273401968122218 -5 -11 -5 6.12 -5
7273401968122219 92 1001 98 6.12 99
…..
5. Access Method and Sort for Supplied Data:
All data files are ASCII files and have been combined into three work sub-directories of the directory wind to facilitate their use.
All metadata files described above in 4.1 and 4.2 sections and this description are put together into one sub-directory Wind_metadata (files "US_1657stns.lst", “US_1657stns.lst_extended", “exist_metadata_files”, “metadata_NCDC_stn.lst”, “anem_elev_inf”, "wind_errors", “wind_extremes”, "wind_problems.1st_run", "wind_extremes.2d_run.TRUSTED" and "wind_extremes_extended.2d_run.TRUSTED").
1657 files with extension “.wnd1” (two of them, 745203.wnd1 and 744920.wnd1, are empty files) are located in the 48_state1 sub-directory and 1304 files with extension “.wnd2” in the 48_state2 sub-directory.
Additionally, a set of 1305 text files with complete digitized station metadata are available upon request. Part of one of them (file 723150.TXT for Asheville, North Carolina) is shown in Table 1 of the Summary Section. One file (745203.TXT, Milford, Municipal Airport, Utah) has metadata but we do not have valid wind information associated with this wmo6 number. While we do have software to process these files digitally, it is not supplied and these files are not a part of the major suite of this archive.
6. Element Names and Definitions are provided for each file in Section 4 and are not repeated here.
7. Start Date: 19310101
8. Stop Date: 20001231
9. Coverage:
a. Southernmost Latitude: 24N
b. Northernmost Latitude: 50N
c. Westernmost Longitude: 125W
d. Easternmost Longitude: 67W
10. Location: North America
11. Keywords:
a. adjusted near-surface wind
b. wind speed
c. wind direction
d. anemometer
e. 10 meters above the ground
f. extreme winds
12. How to Order Data:
Ask NCDC's Climate Services about costs of obtaining this data set. Phone 828 271-4800; Fax 828-271-4876; e-mail orders@ncdc.noaa.gov
13. Archiving Data Center:
13. Archiving Data Center:
a. Name: National Climatic Data Center/NCDC
Address: Federal Building
151 Patton Ave.
Asheville, NC 28801 5001
Voice Telephone: 828 271-4800
Facsimile Telephone: 828-271-4876
Electronic Mail Address: orders@ncdc.noaa.gov
14. Technical Contact:
Dr. Pavel Ya. Groisman
National Climatic Data Center/NCDC
Address: Federal Building
151 Patton Ave.
Asheville, NC 28801 5001
Voice Telephone: 828-271-4347
Facsimile Telephone: 828-271-4328
Electronic Mail Address: Pasha.Groisman@noaa.gov
15. Known Uncorrected Problems:
(a) We found that the data with several wmo6 identificators can be merged together because they are exactly from the same location or from a site that had a small relocation. The merging has been performed and all composite stations are listed in file “US_1657stns.lst” at the last lines (those after 1664). The same file have alternative IDs with the word “merging” in the line with composite stations starting from 70th position. The presence of merged time series should not be considered as a problem per se. But, we sometimes could not trace any logic in partition of data from single station between different wmo6 station IDs: in some cases originally, different hours of the same day were found in different files.
(b) Several stations with different wmo6 IDs have the same WBAN numbers, which indicates that they also could be merged. We found that all but three of these stations have data from different periods of time often with a lengthy interval of “no data” between these periods and only one couple represents a distinctively different neighbor locations. The list of these stations (grouped in couples with the same WBAN numbers) is below. This list was retrieved from file “US_1657stns.lst_extended”, preserves its structure, and is followed with a short line with commentaries when it is appropriate.
1 690070 FORT ORD/FRITZSCHE& CA -121.77 36.68 41 93217 043186 -8 1960 1993 5
1 724916 FORT ORD/FRITZSCHE CA -121.77 36.68 41 93217 043186 -8 1973 1991 5
OVERLAP. But, sometimes there are slightly different values
1 722223 PENSACOLA REGIONAL& FL -87.18 30.47 37 13899 -6 1990 2000 9 ASOS 19971201
1 722220 PENSACOLA FL -87.20 30.47 36 13899 086997 -6 1 1948 2000 9 ASOS 19971201 OVERLAP. But, sometimes there are slightly different values
1 722635 BORGER TX -99.75 30.50 522 13973 414671 -6 1972 1977 7 ASOS 19961202
1 747400 JUNCTION (AMOS) TX -99.77 30.50 522 13973 -6 1977 2000 7 ASOS 19961202
1 722746 SAFFORD MUNICIPAL AZ -109.63 32.85 968 93084 -7 1972 1973 6 ASOS 19970903
1 722747 SAFFORD (AMOS) AZ -109.68 32.82 950 93084 -7 1978 2000 6 ASOS 19970903
1 723139 HATTERAS (ASOS) NC -75.62 35.22 3 93729 -5 1996 2000 9 ASOS 19950701
1 723040 CAPE HATTERAS NC -75.55 35.27 3 93729 311458 -5 1 1957 2000 9 ASOS 19950701 OVERLAP. But, sometimes slightly different values
1 723193 WINSTON-SALEM/SMITH NC -80.22 36.13 296 93807 -5 1990 2000 9 ASOS 19981202
1 723190 WINSTON-SALEM/SMITH NC -80.23 36.13 298 93807 319539 -5 1972 1985 9 ASOS 19981202
1 724243 LONDON-CORBIN ARPT& KY -84.08 37.08 369 03849 -5 1991 2000 8 ASOS 19960918
1 723290 LONDON KY -84.07 37.08 369 03849 154898 -5 1972 1985 8 ASOS 19960918
1 724360 COLUMBUS/BAKALAR IN -85.88 39.25 201 13803 -5 1984 1984 8
1 724363 COLUMBUS/BALKALAR & IN -85.90 39.27 200 13803 -5 1990 2000 8
1 724370 TERRE HAUTE/HULMAN IN -87.30 39.45 181 93823 -5 1972 1985 8
1 724373 TERRE HAUTE/HULMAN& IN -87.32 39.45 178 93823 -5 1990 2000 8
1 725118 HARRISBURG/CAPITAL& PA -76.85 40.22 106 14751 -5 1948 2000 4 ASOS 20001011
1 725110 HARRISBURG/CAPITAL PA -76.85 40.22 106 14751 363699 -5 1972 1985 4 ASOS 20001011 OVERLAP. Values are very close and mostly supplement each other
1 725896 LAKEVIEW OR -120.35 42.22 1455 24270 -8 1977 1997 1
1 725891 LAKEVIEW OR -120.35 42.18 1455 24270 -8 1976 1977 1
1 726459 STURGEON BAY & WI -87.32 44.78 176 04824 -6 1981 1996 3
1 726458 STURGEON BAY WI -87.42 44.85 221 04824 -6 1981 2000 3
These are two different stations with distinctively different metadata.
1 726460 WAUSAU/ALEXANDER WI -89.62 44.92 367 14897 478968 -6 1972 1985 3 ASOS 20000914
1 726463 WAUSAU MUNICIPAL & WI -89.63 44.93 366 14897 -6 1990 2000 3 ASOS 20000914
1 726954 NEWPORT OR -124.07 44.58 48 24272 -8 1990 2000 1
1 726951 NEWPORT OR -124.05 44.63 48 24272 -8 1976 1977 1
1 727927 OCEAN SHORES (CGS) WA -124.13 46.97 3 94225 -8 1972 1990 1 ASOS 20010322
1 727923 HOQUIAM/BOWERMAN WA -123.93 46.97 4 94225 453807 -8 1991 2000 1 ASOS 20010322
1 744200 ROOSEVELT & UT -109.98 40.30 1556 24084 427395 -7 1985 1997 6
1 725708 ROOSEVELT UT -109.98 40.30 1556 24084 -7 1972 1985 6
1 745200 MILFORD UT -113.02 38.42 1533 23176 -7 1983 1985 6 ASOS 19960801
1 724797 MILFORD UT -113.00 38.35 1508 23176 425654 -7 1997 2000 6 ASOS 19960801
(c) The data set has more than 7000 hourly wind speed values that are flagged as grossly erroneous. The user is strongly recommended not to neglect these flags.
(d) ASOS issues. An important homogeneity issue, which has not yet been resolved, was an assessment of the changes in extreme wind speeds with transition to ASOS. This issue is additionally complicated with an algorithm change (so-called firmware 3.0 introduction). This firmware was gradually introduced since 1995. Both, ASOS implementation and firmware replacement directly affect the averaging period used in the hourly wind speed evaluation and, while being irrelevant for mean wind speeds (McKee et al. 2000), changes extreme wind speed estimates. We found that the frequencies of calm weather and extreme wind speed were significantly affected after the ASOS implementation. We also found that the ASOS stations prior to the firmware change (the exact date of which is not always well known) behave quite differently compared to the post-firmware change period as well as with the same period but at the non-ASOS stations. Thus we recommend excluding these data from any possible change assessments that involve extreme and low wind analyses.
(e) Prior to winter 1948-1949, the snow on the ground information from U.S. cooperative stations is sparse everywhere except the Midwestern U.S. and, therefore, we often had to use long-term snow climatology in our formulae (described in Summary) instead of actual daily snow depth information to derive the 10-meter wind speed (variables Sp10 and Sp_surf).
(f) When the day of the anemometer height change was not known (but month and year were shown), it was assumed to be the 15th of this month. Furthermore, in each day of the anemometer height change at each station, we assumed that the change has occurred at 00 LST and used the new height in adjustments of the wind speed to 10 m above the ground/surface.
16. Quality Statement: This data set has undergone extensive quality checks on all parameters, including range checks for wind direction and extreme winds assessment. Questionable data have been flagged.
17. Revision Date: 20011206
18. Source Data Sets: TD-3280, DATSAV3, Metadata archives of the U.S. National Climatic Data Center, U.S. Air Force Combat Climatology Center, and the U.S. Navy Fleet Numerical Meteorology and Oceanography Detachment in Asheville, North Carolina and of the U.S. National Weather Service, Washington, D.C.
19. Essential Companion Data Sets: None. But, acquaintance with documentation in TD-3280 and TD-3505 would be helpful.
20. Derived Data Sets:
A group of derivative climatologies and derived time series have been generated using this data set and its predecessor data set ( in .file2 format) as a source. Among them are:
- mean daily wind time series;
- wind roses;
- monthly/seasonal wind climatology for major wind directions;
- wind climatology during various precipitation events;
- frequency of calm winds;
- frequency of strong winds above several thresholds;
- normalized wind speed distribution for each station and state for the assessment of their changes after the Automated Surface Observation System (ASOS) introduction;
- climatology of “wilting” winds (i.e., winds that occur with surface air temperature above 10?C and relative humidity below 30%); and
- climatology of wind chill (Quayle and Steadman 1998).
Contact Dr. Pavel Ya. Groisman (technical contact) for details how to get these data sets.
21. References:
Court, A. 1953: Wind extremes as design factors. J. Franklin. Inst., 256, No.1, 39-55.
DeGaetano, A.T., 1998: Identification and implications of biases in U.S. wind observation, archival and summarization methods, Theoretical and Applied Climatology, 60, 151-162.
Golubev, V.S., and Bogdanova, E.G. 1996: Accounting of blowing snow events in precipitation measurement in Russia. Proc. of the ACSYS Solid Precipitation Climatology Project Workshop, Reston, VA, USA, 12-15 September 1995, World Meteorol. Organ., WMO/TD No. 739.
Groisman, P.Ya., D.R. Easterling, R.G. Quayle, V.S. Golubev, A.N. Krenke, and A.Yu. Mikhailov, 1996: Reducing biases in estimates of precipitation over the United States: phase three adjustments. J. Geophys. Res., 101, 7185-7195.
Groisman, P.Ya., Easterling, D. R., Quayle, R. G., and Golubev, V.S., 1998: Response to the Comments by D. Yang and B.E. Goodison on the paper by P.Y. Groisman et al. “Reducing biases in estimates of precipitation over the United States: Phase 3 adjustments”. J. Geophys. Res.,103, 6229-6235.
Groisman, P.Ya., R.W. Knight , and T.R. Karl, 2001: Heavy precipitation and high streamflow in the contiguous United States: Trends in the 20th century. Bull. Amer. Meteorol. Soc., 82, 219-246.
Lockhart, T.J. 2000: Summary of wind climate continuity with ASOS. Proc. of the AMS 12th Conf. On Appl. Climatology, 8-11 May, 2000, Asheville, NC, 131-133.
McKee, T.B., N.J. Doesken, Ch. A. Davey, and R.A. Pielke, Sr., 2000: Climate Data Continuity with ASOS. Report for period April 1996 through June 2000. Colorado Climate Center, Fort Collins, Colorado, 25 pp.
Quayle, R.G., and R.G. Steadman, 1998: The Steadman wind chill: An improvement over present scales. Weather and Forecasting, 13, 1187-1193.
22. Summary:
During the NOAA OGP-funded project “Homogeneous blended wind data over the contiguous United States” we have compiled an hourly wind data set from 1655 stations over the lower 48 states for the period of record (up to year 2000), collected and digitized the station metadata related to the history of the anemometer elevation at most of these stations, and blended the station records with historical snow on the ground information. All this has allowed us to generate homogeneous time series of wind speed at the 10-meter height above the ground/surface for the entire period of digital record (usually, since 1948 for the First Order Stations, since early 1970s for other sites with complete metadata, and since the ASOS implementation at the sites without metadata).
We noted that anemometer elevations throughout the U.S. stations varied widely with time. During the past 60 years, there were up to 12 significant anemometer height changes at some of these stations, and on average there was one change per decade at any station with more than 10 years of record. For example, at Los Angeles International Airport the anemometer height changed 4 times during the past sixty years, varying from 59 ft to 20 ft, while at Edwards Air Force Base, the anemometer height was changed 10 times and varied from 13 to 75 ft. Therefore, the elevation homogenization of the near-surface wind time series is a necessary pre-requisite for any climatological assessments.
In this process we were using the formulae:
U10g = Ua log[(10-Hsnod)/z0]/log[(Ha – Hsnod)/z0], and
U10s = Ua log[10/z0]/log[(Ha – Hsnod)/z0],
where z0 is the surface roughness (a function of the presence of snow cover at the site);
Hsnod is the snow depth; Ha is the anemometer height above the ground;
Ua is the wind speed at the anemometer height, U10g is the speed at 10 meters above the ground, and U10s is the speed at 10 meters above the surface.
Left panel of Figure 1 shows the map of those stations for which we were able to perform the homogenization for the period of record. Right panel of Figure 1 shows the stations, for which the homogenization for the pre-ASOS era was impossible due to the absence of metadata (blue dots) and those stations for which any instrument metadata are absent in the NCDC archives. Figure 2 provides general information about the data availability in the archive. Table 1 shows an example of the metadata created for one of the stations.
Figure 1. Left. Stations over the contiguous United States for which complete metadata with anemometer information are available (and thus have been digitized) at the National Climatic Data Center, U.S. Air Force Combat Climatology Center, and the U.S. Navy Fleet Numerical Meteorology and Oceanography Detachment in Asheville, North Carolina. Right. Stations over the contiguous U.S. for which wind information is available in the NCDC data archives in digital form but metadata information with station history is absent. Blue dots indicate the ASOS stations among these stations on both maps.
Quality control issues. For each state we were originally determined to find at least 10 highest wind speed hourly values that can be used as a benchmark for further quality control of these and forthcoming data as well as for practical use. For this purpose for each state, we selected the ten maximum wind speed values (at the anemometer height) and carefully analyzed them by comparing with the entire suite of weather information during the entire day at the neighboring stations including the wind extreme location. Trajectories of Atlantic hurricanes were also used during this analysis. When we saw that the extreme value is unbelievably high compared to those during the neighboring hours, at the neighboring sites, and under the weather conditions in the area (atmospheric pressure and temperature gradients, cloudiness, weather codes, and presence of hurricane/tropical storm/depression nearby were considered), we flagged this value as "questionable" (flag "Q") and continued our search for extremes. Quite often during this analysis, we found that the value most probably has a wrong scale (i.e., multiplied by 10) and scaling back will return it to a proper value similar to those in the neighborhood. We flagged these values as "with scale-error" (flag "S") and also continued our search for extremes. During this analysis we also looked for pattern of errors and tried to reveal them within the entire station (group of stations) records. It appears that no one among the first "extremes" was a true value but a result of some or another type of error. We intended to identify at least 490 extreme wind speed values (10 for each state within the contiguous U.S. but 20 for New Hampshire, where Mt. Washington station represents a group of its own with extremely high winds throughout the entire period of observation) and assign them one of two flags: (1) valid extreme well supported by the neighbor information “E” and (2) valid extreme somewhat supported by the neighbor and weather information “U”. But, prior to identification, we found approximately 7,000 individual wind speed values that were flagged as “Q” or "questionable" and/or “S” or "with scale-error". Furthermore, we checked all wind speed values that were above 30 m sec-1 throughout this process because we believe that whatever the maximum is, the speeds above 30 m sec-1 are too damaging and should be also checked. We also had to exclude 16 pieces of hourly information (from week to up to three years of data) from the station data that have multiple errors of unknown origin (e.g., every second day some of hourly records were replaced with very high wind speed, let say, 30 m/sec, from a given direction (always the same) that is accompanied with negative temperatures (even in summer). These types of errors were discovered mostly on unmanned stations. About eighty wrong extreme wind speed values were discovered in the data originated from TD3280, while the bulk of others came from the DATSAV3 data stream. This was expected because of the highest standards of quality assessment used in the TD3280 data stream. However, it also should be noted that TD3280 represent only ~15% of the stations (also they provide the longest time series) in our data collection. Quality flags are an integral part of each individual wind speed value in the archive.
Figure 2. Annual variations of the number of stations with valid data in the archive and the mean percent of valid hourly data available at these stations. A station was considered present in a given year when it has at least 300 valid hourly wind speed values. Stations that do not have at least 1000 valid hourly records during the entire 1931-2000 period were skipped in this graph (90 of them, including 37 stations with metadata).
Table 1. An example of the wind metadata digitized for one of the U.S. stations, Asheville, North Carolina. Software that processes these tables verifies the ID section of the station information, treats the Y2K and Y19th century problems, and then analyses each record with anemometer elevation information. The days, when this elevation has changed, are in the Remark section; thus this table informs that the first change of the anemometer height (when it was decreased to 20 feet) occurred on September 1st in 1964. ËÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍÍÍ"
º STATION NO. ³ STATION NAME ³ LATITUDE ³ LONGITUDE ³ FLD ELEV (FT) ³ CALL SIGN ³ WMO NMBR º
º ON SUMMARY ³ ³ ³ ³ ³ ³ º
º 723150 ³ ASHEVILLE, NC ³ N 35 26 ³ W 82 33 ³ 2161 ³ AVL ³ º
×ÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄĶ
º STATION LOCATION AND INSTRUMENTATION HISTORY º
ºÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄĶ
º Nmbr ³ ³ Type ³ At This Location ³ ³ ³ Elev Above MSL ³ OBS º
º of ³ GEOGRAPHICAL LOCATION & NAME ³ of ÅÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÅ LATITUDE ³ LONGITUDE ÅÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÅ PER º
º Loc ³ ³ Station ³ From ³ To ³ ³ ³FLD (FT)³BARO (FT)³ DAY º
×ÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄ×
º 1 ³ Asheville (City Office), NC ³ WBO ³ Unknown³ Jan 31 ³ N 35 36 ³ W 82 32 ³ 2253 ³ Unknown ³ º
º 2 ³ Same ³ Same ³ Jan 31 ³ Aug 43 ³ Same ³ Same ³ Same ³ 2266 ³ º
º 3 ³ Same ³ Same ³ Aug 43 ³ Sep 52 ³ Same ³ Same ³ Same ³ Same ³ º
º 4 ³ Asheville (Airport), NC ³ WBAS ³ Sep 52 ³ Sep 64 ³ N 35 26 ³ W 82 32 ³ 2161 ³ 2161 ³ º
º 5 ³ Same ³ Same ³ Sep 64 ³ Dec 64 ³ Same ³ Same ³ Same ³ Same ³ º
º 6 ³ Same ³ WSO ³ Dec 64 ³ Sep 74 ³ Same ³ Same ³ Same ³ Same ³ º
º 7 ³ Same ³ Same ³ Sep 74 ³ Mar 87 ³ Same ³ Same ³ Same ³ Same ³ º
º 8 ³ Same ³ Same ³ Mar 87 ³ Nov 89 ³ Same ³ Same ³ Same ³ Same ³ º
º 9 ³ Same ³ Same ³ Nov 89 ³ Jun 96 ³ Same ³ Same ³ Same ³ Same ³ º
º ³ ³ ³ ³ ³ ³ ³ ³ ³ º
ÎÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍØÍÍØÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÎ
º Nmbr ³ Date of ³ SURFACE WIND EQUIPMENT INFORMATION ³ º
º of ³ Change ³ÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÅ REMARKS, ADDITIONAL EQUIP- º
º Loc ³ ³ LOCATION ³ TYPE OF ³ TYPE OF ³ HT ABOVE ³ MENT, OR REASON FOR CHANGE º
º ³ ³ ³ TRANSMITTER ³ RECORDER ³ GROUND ³ º
×ÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄĺ
º 1 ³ Jan 31 ³ on roof ³ F102 ³ Unknown ³ 92 ft ³ 01 F011 indicator º
º 2 ³ Aug 43 ³ Same ³ Same ³ Same ³ Same ³ º
º 3 ³ Sep 52 ³ Same ³ Same ³ Same ³ Same ³ º
º 4 ³ Sep 64 ³ 497' E of admin bldg ³ F420D ³ Same ³ 20 ft ³ 01 º
º 5 ³ Dec 64 ³ Same ³ Same ³ Same ³ Same ³ º
º 6 ³ Sep 74 ³ Same ³ F420A ³ F312 ³ Same ³ 01 º
º 7 ³ Mar 87 ³ Same ³ Same ³ Same ³ Same ³ º
º 8 ³ Nov 89 ³ 720'WSW of admin bldg ³ F420R ³ Unknown ³ Same ³ 01 º
º 9 ³ Jun 96 ³ Unknown ³ S100 ³ S100 ³ 26 ft ³ 01 ASOS/ records end º
º ³ ³ ³ ³ ³ ³ º
ÎÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍØÍÍØÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÍÍÍØÍÍÍÍÍÍÎ
º Nmbr ³ Date of ³ PRECIPITATION GAGE INFORMATION ³ º
º of ³ Change ³ÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÅÄÄÄÄÄÄÄÄÄÄÅ REMARKS, ADDITIONAL EQUIP- º
º Loc ³ ³ LOCATION ³ TYPE OF ³ WIND ³ HT ABOVE ³ MENT, OR REASON FOR CHANGE º
º ³ ³ ³ GAGE ³ SHIELD ³ GROUND ³ º
………………….
1 If the station became Automatic Weather Observing station (AWOS, usually around 1990) and later was converted to ASOS site, then only ASOS type is indicated by a four-character abbreviation “ASOS” or “ASOs”.
2 This 1-character code indicates which cloud element from TD 3280 and/or DATSAV3 was used for cloud information and can be blank (for CLCx), 1 for TSCE, 3 for ALCx, 5 for ALMx, and 8 for TSKC. See description of TD 3280 for meaning of CLCx, ALCx, ALMx, TSCE and TSKC. These variables are beyond the scope of the present archive.
3 Quality flag can be: blank (no adverse comments), ‘Q’ – questionable value that is most probably wrong and cannot be repaired, ‘S’ – questionable value that is most probably wrong but a scale factor 0.1 brings it in coordination with neighbor hours and sites; ‘E’ – valid extreme value that is supported by a widespread evidence from neighbor hours and sites and synoptic information (for example, value of 30.3 m sec-1 on October 5th 1948 at 6 PM in Miami has this flag); ‘U’ - valid local extreme value that has some limited support from neighbor hours/sites and/or weather information (e.g., present weather code reports squall, blowing dust/snow, or heavy thunderstorm); ‘H’ – indicates a presence of hurricane with the center in the 300-km neighborhood of the site (as in Miami in October 5th 1948); and ‘T’ – indicates a presence of tropical storm in the vicinity of the site (in radii of 150 km).
17