Saturday, March 7, 2020

KDCA, KBWI, and KIAD Climatological Winter 2019-2020 Review: PART 1: A Meandering Bit of Historical Background Info; PART 2: Overview - Warm, No Snow, and No Sterling Quality Control

**Updated 2:34 a.m. 3/25/2020: See below.**

So, with no public information statement or, for that matter, any notice, the National Weather Service (NWS) Forecast Office Baltimore / Washington (i.e., "Sterling" since it is located in Sterling, Va., or simply "LWX") updated the monthly average temperature files for the three airport climate stations earlier today, specifically, the February 2020 average temperatures for KDCA, KBWI, and KIAD.

I refer to the files containing the PDFs of temperature data back to the dawn of official weather-keeping at the three spots -- and in the case of KDCA and KBWI, their predecessor spots.

For what is now "KDCA" and "KBWI," that would be all the way back to Nov 1, 1870, the very dawn of official U.S. Government weather record-keeping started.

Part 1: Historical Context

At this point, I'm going to take a meandering detour into the history of U.S. weather observations starting with an image of several paragraphs from this detailed NWS/NOAA document (that is an FTP link and may result, depending on your browser, in the 46MB PDF document being downloaded).

Nine months earlier on February 2, 1870 (yes, coincidentally on what later became Groundhog Day) that Congress uneventfully passed a Joint Congressional Resolution authorizing what is known today as the U.S. Army Signal Corps, then housed in the War Department, to undertake weather observations. President Grant signed it into law on February 9, 1870.

There were 24 stations in the initial set to include one station in Washington, D.C., and one Baltimore, Md. -- and none of which were at airports since, of course, airports didn't then exist (except maybe in Donald Trump's deranged mind).

Screenshot of Sterling's Unique Local Climate Data web index page; direct link here.

Just change the "lwx" in the URL to any other three-letter NWSFO office call sign (i.e., identifier) to get its corresponding page. For instance, "lox" for Los Angeles, "okx" for New York City, and "lot" for Chicago.


In terms of the historical "climate record" available on the NWS Forecast Office sites, Jan 1, 1871 is more typically listed as the start date for at least basic parameters such as temperature, barometric pressure, wind, and rainfall. Snowfall measurements and dew point came at later points.

And, to be clear, earlier records exist back into the Colonial era. It's well known that George Washington, Thomas Jefferson, and, yes, Benjamin Franklin kept weather observations with Franklin even recording conditions in Philadelphia on July 4, 1776 (see image at left).

There are also sets of systematic records pre-dating 1871 in cities such as New York and San Francisco.

With all that in mind …

The information below is taken from two sources:

A 36-page document History of Weather Observing in Washington, D.C., 1821 - 1850 by Gary K. Grice (prepared in Feb 2005) and a very old book Maryland Weather Service. Reports. New Series. Volume 2 (image of cover shown left).

The latter was produced by U.S. Weather Bureau meteorologist and climatologist Oliver Lanard Fassig published in 1907.

Dr. Fassig was quite an interesting fellow in his own right. Among his achievements, he was the first person to ever earn a PhD in meteorology in the United States (at Johns Hopkins University).

While he lived to 76-years old, he met an unfortunate end in that he was struck by an automobile. He died on Dec 6, 1936.

Anyway, back to November 1870 and the start of formal U.S. Government weather record-keeping …

Here in D.C., record-keeping started at Signal Service HQ at 1719 G Street NW (pictured left in the 1890s) and in 1888 moved nearby to 1744 G Street NW. In 1889, record-keeping moved to Signal Service / Weather Bureau Office (pictured below circa 1916) at 2416 M Street NW.

Thereafter, in the "aviation era," it moved to what is now Joint Base Anacostia - Bolling (between Aug 1929 and July 1931), then to the old Washington Hoover Airport (July 1931 - June 1941) and, thereafter, to present-day National Airport, although not at the exact spot as today.

Map of locations of early Signal Corps and Weather Bureau weather stations in Washington, D.C.
(using a contemporary map)


However, records were sometimes kept in parallel such as during the 1940s when records were also kept at the reconstituted Weather Bureau Office at 2400 M Street (the Signal Corps having been moved into the U.S. Army after the 1942 organization that gave us the Department of Defense). The above dates refer to what were the sites of the "official" record-keeping. What is considered "official" is simply decreed by the NWS.

Signal Service and Weather Bureau Office headquarters, Washington, D.C., at 2416 M Street NW, circa 1916


The move to National Airport occurred on June 16, 1941 right when the airport opened, but the "official" record period for KDCA is considered to have started in August 1945. The spot was atop the old A terminal control tower (see image below).

National Airport's old A terminal original control tower and its rooftop weather station pictured circa 1945.


This remained the spot until August 30, 1968 -- presumably, at which point, it moved to its current spot in the far southeast corner of the airport field by the Potomac in what is the worst location possible in the entire region for measuring cold temperatures and snowfall.

Map of Washington, D.C., "aviation era" weather stations between 1929 - 1968


In the case of Baltimore, there were five locations for official weather record-keeping sites … The very first was the Fireman's Insurance Building at the southwest corner of South and Water Streets for 14 years … second, for 2-1/2 years in Neal Office Building at the southwest corner of Holliday and Baltimore Streets … third, for a year at the Johns Hopkins University (JHU) Physical Laboratory … fourth, the Equitable Building at the southwest corner of Calvert and Fayette Streets … and then back to JHU in the Treasurer's Building at 532 Howard Street.

Screenshot of page 304 of the Fassig book.

I would note that page 305 gives a date for the first building being "occupied" as Dec 23, 1870. Even that doesn't agree with the Nov 1, 1870 date -- but I'm not going to try to resolve matters at that level of granularity.

Eventually, the weather observations ended up at the U.S Custom House -- although there were two intermediate airport weather stations with observations in parallel to the U.S. Custom House, namely, Logan Field, a.k.a., Dundalk Flying Field, in Dundalk, and Baltimore Municipal Airport, a.k.a., Harbor Field, and today the location of the Dundalk Marine Terminal.

Finally, it was moved in July 1950 to the then-brand new, then-Friendship Airport (pictured above circa 1960), now BWI Airport, although record-keeping at the Custom House continued until around the time the Inner Harbor ASOS at the Maryland Science Center ("KDMH") came online in April 30, 1998. (The "temporary" KDMH shutdown due to construction around the ASOS site has yet to start.) For KIAD, that would be January 1960, although continuous records there don't start until Jan 1964.

Part 2: Overview of This Winter

At this point, let's return to the original point of this entry, namely, an overview of the climatological winter of 2019 - 2020.

It was last week that Sterling updated the three web-based PDF files containing February 2020 and winter 2019-2020 temperature information. The index page for these is available here, and I refer to the following three links under "Climate Data" [I've embedded the links in each one, but depending on the browser you use, it will either open up the page or just download the PDF]:

Washington Average Monthly Temperature (since 1871)

Baltimore Average Monthly Temperature (since 1871)

Dulles Average Monthly Temperature (since 1960)

Also updated was the "WINTER" column -- to refer to the period of climatological winter of 2019-2020, i.e., Dec 1, 2019 - Feb 29, 2020), rather than astronomical or "seasonal" winter. It shows up on the row for 2020.

The issue is that the numbers make no sense, especially for KDCA. The pages show the following monthly anomalies, which except for a 0.1F temp difference in the December normal of 39.7F versus 39.8F, matches the CF-6 preliminary data linked from here):

Dec 42.2F (+2.5F) / Jan 42.4F (+6.4F) / Feb 43.8F (+4.8F)

Dec 39.9F / (+3.2F) Jan 40.2F (+7.3F) / Feb 42.0F (+6.2F)

Dec 39.0F (+2.4F) / Jan 38.9F (+5.7F) / Feb 41.4F (+5.2F)

Using a simple arithmetic mean, we obtain the following (with averages in parentheses):

KDCA: 42.8F or +4.6F (38.2F)
KBWI: 40.7F or +5.6F (35.1F)
KIAD: 39.8F or +4.4F (35.3F)

I ran the numbers myself -- see above image of a simple Excel spreadsheet that I made -- and you'll notice in the case of KIAD, there is a 0.1F discrepancy -- it should be +4.5F if the average is, in fact, 35.3F. However, the issue is resolved in you do a fractional adjustment to account for the fact that December and January were 31 days and February was 29 days, it resolves itself. Specifically, for the winter 2019 - 2020, KIAD had an average temp of 39.7F or +4.4F above average.

Not surprisingly -- since Sterling never seems to quality control its numbers -- it came up weirdly different values (discrepancies in parentheses):

KDCA: 43.2F (+0.4F)
KBWI: 40.6F (-0.1F)
KIAD: 39.7F (spot on)

Just FYI, the KDCA numbers that I calculate match what appears in this CWG entry (link embedded):
Snowless and abnormally mild February caps Washington’s seventh-warmest winter on record.

I also got a few images in this entry from that posting.

Using these numbers and comparing them to the rest of the full monthly average files, we can rank the winters as follows:

KDCA: 7th warmest
KBWI: tied for 6th warmest
KIAD: tied for 3rd warmest

In the case of KDCA, the top 7 ranked warmest winters are as follows:

1931-'32: 44.6F
1889-'90: 44.3F
2016-'17: 43.9F
2012-'12: 43.4F
2001-'02: 43.2F
1949-'50: 42.9F
2019-'20: 42.8F

Two of the seven are in the pre-airport period, and of the other five, they're all since 2000.

In the case of KBWI, the top 6 ranked warmest winters are as follows:

1931-'32: 45.3F
1889-'90: 44.4F
1948-'49: 42.3F
1949-'50: 42.0F
1871-'80: 41.3F
2011-'12 / 2019-'20: 40.7F

Two of these seven winters are in the airport period (2011-'12 and 2019-'20) while the other five are not.

This is a reflection of the fact of how much warmer it is and was in downtown Baltimore. That is, even in our globally warmed era, it is still not a good comparison.

In the case of KIAD, the warmest winters are:

2016-'17: 40.8F
2011-'12: 40.1F
2001-'02 / 2019-'20: 39.8F

For some additional information per two Sterling (LWX) tweets:

The lowest temp during meteorological winter 2019-'20 was only 22F at KDCA, 19F at KBWI and 15F at KIAD.

All of these set records for being the warmest minimum temperatures ever recorded for met winter (for KBWI, this only counts the airport period since downtown temps have stayed warmer).

The lowest HIGH temp during meteorological winter 2019-'20 was only 35F at KDCA, 33F at KBWI and 34F at KIAD. All of these set or tied records for being the warmest of the coldest high temps ever recorded for met winter.

Updated 2:34 a.m. 3/25/2020 

OK, I should note that Sterling (LWX) corrected its numbers and now its numbers on the PDF pages match the ones above, although KIAD at 39.8F, not 39.7F.

This occurred after an email exchange with them. I would like to note this fact.

Also, my apologies for above commentary, too.

End of Update.


Snowfall …

Needless to say, snowfall was practically non-existent this winter.

Through Feb 29, the totals, departures / averages for the season since July 1st, and full seasonal averages are as follows

KDCA: 0.6" -13.5" (14.1") [15.4"]
KBWI: 1.8" -16.4" (18.2") [20.1"]
KIAD: 2.9" -16.0" (18.9") [22.0"]

The Feb totals were trace at KDCA and KIAD and 0.0" at KBWI.

Keep in mind that the new NWS normal base period of 1991 - 2020 is going to come out next year and the snow totals will drop a bit -- even while the winter temps go up a bit. Of course, with Sterling, the numbers could be whatever …

One thing that won't change as long as one individual continues to oversee the gaggle of FAA contractors who take the weather observations at National Airport -- he's been doing it since 1979, which is far too frickin' long for one person to do it precisely for reasons of bias -- is the apparently "inability" to surpass the 1979 Presidents Day snowstorm total, which is ranked as #1 in the KDCA period and #3 overall.

By contrast, KBWI and KIAD have each managed to surpass that total multiple times since then.

OK, that's all. Finally, I got this damn entry completed.


No comments: