Formats and abbreviations used in my USHCN station comparison summaries, (using Fort Dodge, and Iowa Falls, Iowa as an example). Some of the following is extracted, or adapted, from a USHCN readme file such as: ftp://ftp.ncdc.noaa.gov/pub/data/ushcn/README.TXT Lines from the example are prepended with > >Station# Lat Long Altd ST Name >132999 42.50 -94.20 1115 IA FORT DODGE >134142 42.52 -93.25 1130 IA IOWA FALLS Station#: station number/identifier : 2 digit State Code (01 Alabama through 48 Wyoming) : 4 digit Coop Station Number Lat : latitude in degrees and fractions of degrees (not minutes) Long : longitude in degrees and fractions of degrees (not minutes) Altd : "current" altitude above sea level in feet circa 1994 ST : state abbeviation Name : "current" station name circa 1994 Next will be a "time of observation" line, followed by associated station history entries. > 134142 obsv time 07:00 132999 obsv time 07:00 > 134142 03 01 1892 obtim 999 999 1107 0025 E 9079 00CD000000 04 J B PARMELEE > 132999 04 01 1904 obtim 000 000 1126 9999 999 9079 00CD000000 05 MONK & FINDLEY Starting times of observation for period of the summary, followed by extracts of USHCN station history file of the associated entries. The history extract for station two is indented two extra spaces. Subsequent changes to the time of observation at either station will include the date, and a code indicating whether other changes occurred concurrently. The extracts from the station history file include: station number month day and year of the entry my code for kind of change(s) displacement from previous location direction of that displacement altitude above sea level in feet distance from post office direction of that distance time of observation code equipment codes height (in feet) of temperature sensor above ground name of observer A note of caution: station changes after 1994 appaar not to have been included in the USHCN station history file as of the year 2000 update. Descriptions of the various codes, and units, are further below. Next are two header records for the following temperature data: >Annual temperature means and adjustments and some averages thereof >Year Stn_1 R_1 TA1 FA1 F_1 Stn_2 R_2 TA2 FA2 F_2 R1-2 F1-2 year of data Stn_1 station number of the first of the pair of stations in the summary R_1 "raw" annual mean temperature for station one in degrees F TA1 USHCN time of observation bias (TOB) adjustment FA1 USHCN station change (SHAP,FILNET), and MMTS, adjustments F_1 USHCN adjusted temperature (i.e. "raw" plus net adjustments) Stn_2 station number of the second of the pair of stations in the summary R_2 "raw" annual mean temperature for station two in degrees F TA2 USHCN time of observation bias adjustment FA2 USHCN station change (SHAP,FILNET), and MMTS, adjustments F_2 USHCN adjusted temperature (i.e. "raw" plus net adjustments) R1-2 difference between the "raw" annual temperatures of the two stations F1-2 difference between the adjusted temperatures of the two stations I use the term "raw" to refer to the USHCN "areal edited" temperature data prior to TOB adjustments. The Areal Edited data is the original (or raw) data that have been screened to flag monthly data that are suspect/outliers (over 3 standard deviations from period of record mean of the element). A sample as described above with a few years of data. Missing years did not have a "raw" annual temperature in the USHCN mean temperatures file for at least one of the two stations. All of these USHCN temperature data are in degrees Fahrenheit. >Annual temperature means and adjustments and some averages thereof >Year Stn 1 R 1 TA1 FA1 F 1 Stn 2 R 2 TA2 FA2 F 2 R1-2 F1-2 >1905 132999 45.70 0.50 -0.38 45.82 134142 44.70 0.49 0.92 46.11 1.00 -0.29 >1906 132999 46.80 0.45 -0.38 46.87 134142 46.20 0.44 0.91 47.55 0.60 -0.68 >1908 132999 47.80 0.48 -0.38 47.90 134142 47.10 0.49 0.92 48.51 0.70 -0.61 after sets of years without temperature related station change entries, averages will follow: >Avg 11 yrs: 46.32 0.47 -0.38 46.40 44.98 0.46 0.91 46.35 1.34 0.05 >Net adjustments: \ 0.08 / \ 1.37 / \-1.29/ >Ranges: 5.38/\ 2.80 5.75/\ 3.74 1.51/\ 1.36 The net adjustments line displays differences between the indicated averages. The ranges line displays three, or three pairs, of ranges of the R 1 ("raw" annual means of station one), R 2 ("raw" annual means of station two), and the R1-2 (R 2 subtracted from R 1), columns. If the number of years being averaged is less than four, one range (max - min) of each of those three columns will be displayed. If the number of years being averaged is more than three, a second range for each of those three columns will be displayed. Each second range is the max - min of the years being averaged excluding those of the max and min of the first range for the corresponding column. The USHCN "raw" mean monthly, and mean annual, temperatures are rounded to tenths of a degree F, but the adjusted mean monthly, and mean annual, temperatures are not. So, I use a "raw" mean annual temperature recalculated from the monthly means, instead of the "raw" mean annual temperature as published. The adjustments that I call "station change" adjustments are those other than the TOB adjustments. They seem usually to be in response to station moves, equipment changes and/or observer changes. Occasionally, they will reflect a special adjustment of a single month due to some unusual aspect of that month's mean temperature at that station, such as that the data were in excess of 5.0 standard deviations from their mean offset with respect to the station's nearest neighbors. I do not include UHI adjustments in these comparisons. Station history entries which indicate no temperature related change are included in these summaries "just in case". They may refer to precipitation related changes, or to something such as distribution of station reports, etc. The extracts from the station history file include the following kinds of codes: My codes for kinds of changes: nomov : appears only in time of observation line, and indidcates that no other temperature related change was recorded beside the time change. moved : when in a time of observation line, indidcates that a move also occurred, but no other temperature related change was recorded with the time change. When not in a time of observation line, it was the only temperature related change recorded. obtim : station history entry which triggered the preceding time of observation line. eqpch : an equipment change was reported, but not a move or observer change obsch : an observer change was reported, but not a move or equipment change precp : station history entry which indicates a precipitation related change, but is included in the summaries "just in case" other : station history entry which does not indicate a temperature related change, but is included in the summaries "just in case" combinations of codes: mvobc : location move and observer change was reported mveqc : location move and equipment change was reported mvoec : location move and observer change and equipment change was reported When included in time of observation lines, these codes will be preceded by a < to indicate that the change was at station one, or followed by a > to indicate that the change was at station two. Station history entries, excepting those designated "other" or "precp", are treated as break points separating sets of data included in multi-year averages. NOTE: In the context of the USHCN, the location of temperature instrument defines official station location. Displacement (i.e. change of location) codes: DISTANCE FROM PREVIOUS LOCATION (DPL) units in tenths of miles unless followed by a "B"; then the units are in "Blocks"; "999" = unknown distance (NOTE: Distances > 80 miles are flagged as suspect, "9xx" codes indicate distance value is exclusively for the temperature instrument; "8xx" codes indicate distance value exclusively for precipitation.) DIRECTION FOR DPL codes for the 16-points are used: N, NNE, NE, ENE, E, ESE, SE, SSE, S, SSW, SW, WSW, W, WNW, NW, NNW; "999" indicates direction unknown; "000" indicates no change in direction except for some instances of precipitation only moves [see "8xx 000" example below] Examples of DPL Codes: CODE DEFINITION 999 999 Distance & direction unknown (the DPL is always missing in the first record for a new station) 015 NW Station moved 1.5 miles NW of previous location 000 000 No change in either station or instrument location 902 ESE Temperature instrument moved 0.2 miles ESE and precipitation instrument did not move; or the precipitation instrument move was not equal to that of the temperature instrument 800 000 Precipitation instrument moved, but temperature instrument did not move (in more recent history entries the direction may be given rather than encoded as zeroes) 000 ESE Station moved < 0.1 mi ESE of previous location 999 NW Moved an unknown distance to a location NW of the previous location ELEVATION (of ground at temperature site) given in whole feet above or below mean sea level Distance from post office codes: DISTANCE FROM POST OFFICE (DPO) units are in tenths of miles unless followed by a "B"; then units are given in "Blocks"; "9999" = unknown distance DIRECTION FOR DPO codes for the 16-points are used: N, NNE, NE, ENE, E, ESE, SE, SSE, S, SSW, SW, WSW, W, WNW, NW, NNW; "999" indicates direction unknown; "000" indicates a location at the PO Examples of DPO Codes: CODE DEFINITION 9999 999 Distance & direction unknown 0015 NW Station 1.5 Miles NW of PO 0000 NW Station < 0.1 Mile NW of PO 0000 999 Station < 0.1 Mile from PO; direction unknown 9999 NW Station located an unknown distance to the NW of the PO 0000 000 Station located at PO Time of observation codes: The four position field for times of observation may show precipitation observation time in the first two position, and temperature observation time in the second two. However, other possibilities are handled by codes described below. The observation times are encoded to the nearest whole hour where known; in some instances, the observation may have been taken at a variable hour (e.g., sunrise or sunset) or according to a rotating seasonal schedule based upon agricultural needs. Examples: 01-24,99 Nearest Whole Hour; "99" for unknown SR Sunrise SS Sunset RS Rotating Schedule (in general, evening observations were taken during the winter months [Oct/Nov thru Apr/May] and morning observations taken during summer months or "crop season".) In "RS" instances, often two observations were taken during the summer; usually between 5-7 AM and again between 5-7 PM. In such 2 obs/day instances, it's unclear whether both elements are read twice, but recorded for a single hour; or (a more likely practice), the precipitation is recorded in the morning and temperature in the evening. At some sites (usually airports) a uniform schedule was in use, but the particular hour at which a particular element was observed was not specified. In these instances, the number of hours per day for which observations were taken is given. (NOTE: the number of hours must not be confused with synoptic observing schedules such as 3-hourly, 6-hourly, etc. ones at which observations are taken at intervals of every 3 hours, 6 hours, etc.) xxHR where xx indicates the number of hours for which the site took observations; An early practice regarding the observation times of temperature involved taking three readings (morning, afternoon, and evening or "tri-daily"). These readings were usually made at 7 AM, 2 PM, and 9 PM. A weighted average was computed in which the 9 PM observation was summed twice and the total sum divided by 4 [e.g., (7 AM + 2 PM + 9 PM + 9 PM )/ 4 ]. Note, in these instances, no precipitation observation time has been encoded. TRID indicates "tri-daily" temperature average The early editions of the Weather Bureau's reporting forms allowed space only for a single observing time to be entered. Later on, revised forms specifically identified whether the observation time was for temperature or precipitation. In the instances where the forms were ambiguous with regard to the observing times of these elements, the OT is encoded as follows: 9xx9 where xx indicates the hour as listed for the observations; whether it's accurate for both temperature and precipitation is unknown Examples of OT codes: SRSS Precipitation reading made at sunrise; temperature read at sunset SS99 Precipitation read at sunset; temperature read at an unknown hour or there is no temperature data available for that period of record 0718 Precipitation read at 0700 (i.e., 7 AM); temperature read at 1800 (i.e., 6 PM) 9079 Ambiguous form: 0700 was the only listed hour of observation; unknown whether for both temperature and precipitation 06HR Station observed 6 hours/day; whether the average of these 6 observations is what was recorded or only a single hour was used is unclear RSSS Precipitation readings made on a rotating schedule; temperatures read at sunset TRID "Tri-daily" readings of temperature; no OT encoded for precipitation 9918 Precipitation readings made at an unknown hour or there is no precipitation data available for that period of record; temperature read at 1800 Equipment codes: A Cotton Region Shelter (i.e. Stevenson screen) B Dry-Bulb Thermometer C Minimum Thermometer D Maximum Thermometer E Thermograph F Digital Thermometer G Maximum/Minimum Temperature System H Hygrothermometer (type unknown) I Hygrothermometer - H06x series J Hygrothermometer - H08x series K Automated Meteorological Observing System