Canada/U.S. METAR -- Mutations of the International Standards

by Gene A. Nygaard

The United States changed its aviation weather reports to a new format at 0800Z on 1 July 1996. Canada changed to the new format 28 days earlier at 0600Z 3 June 1996. (Some Canadian stations, however, still use the old SAO format.) Both countries call this new format METAR, though the reports bear only a superficial resemblance to the international reports of that name. (Canada's reports differ from the U.S. reports in the remarks section, but the main body of the report is pretty much the same except for a difference in the way temperatures are actually measured.)

For those of you unfamiliar with this, here are some examples of these coded reports. It is beyond the scope of this page to explain them, but some of the links here have more information about them.


METAR KHEI 182353Z AUTO 14007KT 10SM CLR 11/M06 A2957 RMK AO2
SLP031
      6//// T01061056 10133 20100 56022 TSNO PNO

METAR CYQW 190000Z 10013KT 15SM FEW080 SCT100 BKN250 02/M06 A2936
RMK
 AC2AC2CI2 SLP965

WHAT IS THE PURPOSE OF MAKING THIS CHANGE? The way it is being done, very little is being accomplished towards a goal of an international standard for aviation weather reports.

The U.S. METAR code is described in Federal Meteorological Handbook (FMH) No. 1 "Surface Observations and Reports" (Dec 1995).

According to that handbook, there are seven weather elements in the body of a METAR report. These elements are:

These elements are reported in the same order in CANUS METAR as they are reported in the international METAR reports. But beyond this, there is little resemblance once you get past the date/time code.

DIFFERENCES FROM THE INTERNATIONAL STANDARDS

The U.S. National Weather Service has a METAR/TAF Home Page on the Worldwide Web. These pages include an Overview of METAR and TAF. This is what it says about some of these elements of the reports which the U.S. calls by the METAR name [bullets added]:

"For example, in order to lessen the burden on the U.S. aviation community, a number of exceptions to metric reporting units have been filed by the U.S.

Note the improper capitalization of hectopascals also.

Another way in which what the U.S. differs from the international standards is spelled out in another file available at the METAR Home Page which lists U.S. exceptions to the METAR and TAF regulations: "15.10 CAVOK shall not be used."

In other words, the U.S. National Weather Service admits that 5 of the 7 elements differ greatly from the international standards.

Environment Canada also has a TAF/METAR Home Page (the U.S. and Canadian pages also have links to each other).

I also wonder what Canada's excuse is for using the same units--they shouldn't have the same "burden" in a country that is much more metric than the United States is.

TEMPERATURES AND DEW POINTS ARE ALSO DIFFERENT


METAR KHEI 182353Z AUTO 14007KT 10SM CLR 11/M06 A2957 RMK AO2 SLP031
      6//// T01061056 10133
20100 56022 TSNO PNO

That is only part of the story. The Overview continues:

"The only element that will be converted to metric units is the temperature/dewpoint field which will be reported in whole degrees Celsius. In order to allow for
This is the only "official" reason given for adding about 20% to many reports in what is supposed to be a compact, efficient way to transmit information. No one at NWS or FAA claims that this Tttttdddd element is needed for aviation purposes.
a better conversion between Celsius and Fahrenheit, the hourly temperature/dewpoint will be in tenths of degrees Celsius in the additive data remarks section of the METAR report from selected stations in the U.S."

This is not the way it was done in the United States in the past. It is also not the way it is done in any other country in the world (including Canada, which apes the U.S. in most of the rest of its butchering of the METAR standard). What's more, it is not even the way that Federal Meteorological Handbook No. 1 says that it shall be done!

FMH-1 says that it prescribes the standards for observing temperature and dew point, as well as reporting them (10-2). It says that temperature "shall be determined to the nearest tenth of a degree Celsius." (10-4-2, emphasis added.) It says that dew point "shall be determined to the nearest tenth of a degree Celsius." (10-4-3, emphasis added.) It says that maximum and minimum temperatures in a six-hour period "shall be determined to the nearest tenth of a degree Celsius." (10- 4-4, emphasis added.) Note that it is the physical quantities being measured (i.e., temperature, dew point) which SHALL BE DETERMINED, and there is no mention of any conversion from a whole number of degrees Fahrenheit.

But that is not what is being done, except for a few stations.

The temperatures reported in the main body of a U.S. report will be wrong 2 times out of every 15 on average, assuming the original measurements are accurate. The way these are calculated is that the instruments measure temperature and dew point in a whole number of degrees Fahrenheit. This rounded-off number is then converted mathematically to the Celsius scale. For example, if the temperature is 32.7 degrees F, then this will be measured as 33 °F. This is equal to 5/9 of a degree Celsius, which will be rounded to 1 °C for the body of the report. But 7/10 of a Fahrenheit degree above the freezing point is only 7/18 of a Celsius degree above the freezing point. Since 7/18 is less than 9/18, it should be rounded to 0 °C, not the 1 °C which is reported.

The temperatures reported in the remarks section of a U.S. report will be incorrect 82% of the time.
Here's a strange . . .
There is a major difference in the coding of the whole degrees Celsius in the body and the tenths in the remarks. They use different signs for the numbers. In the body, the positive sign is omitted, and the negative sign is "M"; but in the remarks, the positive sign is "0" and the negative sign is "1".
In any 5 Celsius degree interval, the tenths of a degree divisions total 50 numbers. But of these, only the 9 numbers corresponding to conversions from whole numbers of degrees Fahrenheit will be used. What is reported as 4.4 degrees Celsius might actually be 4.2 or 4.3 or 4.5 or 4.6 or 4.7 degrees, not necessarily 4.4.

So this makes a sixth element which differs considerably from the international standards. The remaining element, Present Weather, does use most of the international codes such as RA for rain, FG for fog, and GR for hail. This may be the element which most closely resembles the international standard. (I wouldn't bet very heavily on it being the same in all respects, however.)

Statute miles are an anachronism

The use of statute miles in an aviation context is an anachronism. It is a holdover from the days when most of the air speed indicators in North American aircraft reported the speed in statute miles per hour, and aeronautical charts were drawn at a scale of a whole number of statute miles to the inch. Charts in both countries are drawn on a metric scale (1:500 000, 1:1 000 000), and they use miles which are unidentified (apparently "everybody" already knows what they are, but I had to measure them and use the map scale to determine that they are the 1.852 km variety of miles, not statute miles).

Statute miles do not fit in with anything else in aviation weather. There was no compelling reason to deviate from the international standard of reporting visibilities in metres which is followed by all of the 150-odd countries that were using the METAR format before 1996.

CANADIAN PUPPETS

I'd like to make a couple of additional comments about the Canadian mimicry of U.S. usage in what these two countries call METAR.

It is hard to understand why Canada follows the United States in using statute miles for visibility in the aviation reports, when in the public reports (such as when Environment Canada wants to warn motorists to stay off the roads during a blizzard), the units used are meters, which are the international standard for aviation reports.

In addition, Canada follows the U.S. in telling users to decode sea level pressures (included in the remarks section, not the main body) as hectopascals. (The office in the U.S. which wrote the FMH-1 handbook has known at least since 1991 that these are not proper SI units, but they use them anyway.) Environment Canada does this even though to get the correct SI units, kilopascals, which are the units used in Canadian public weather reports for many years and which will continue to be used in the future in those reports, it would only be a matter of telling users to insert a decimal point before the second to last digit rather than before the last one. (The 1 or 2 most significant digits must also be inserted to get the actual reading in pascals with any prefix.)

International System of Units

Back to Gene's Weather Page

Send your comments to Gene Nygaard,
P.O. Box 591
Crosby, North Dakota 58730
U.S.A.