« March 2019 | Main | May 2019 »

April 2019

National Weather Service's Year-Round Use of 'Standard Time' Can Cause Confusion

Daylight savings and standard time


I had been writing this blog for a number of years before learning that the National Weather Service doesn't recognize Daylight Saving Time (the period roughly between mid-March thru early November).  Instead, its reported observations throughout the year reflect Standard Time.  Because of this policy, weather conditions that occur between midnight and 12:59 AM during DST are reported as part of the previous day's weather observation (credited to 11:00-11:59 PM).  And the rest of each day's weather observations are reported as happening in the previous hour.

I discovered this quirk because I check hourly observations every day on one of the NWS sites, and during the spring, summer and first half of the fall (i.e., when DST is in effect) I noticed that rainfall and temperature for the first hour of each day was credited to the previous day.  Then when Standard Time resumed the weather conditions of each day's first hour were back to being part of that same day.


Questioning man

I thought that, perhaps, this decision was made over concern by the NWS about how to report conditions that occur during the "skipped" hour, i.e., when we "spring ahead" at 2:00 AM for DST (losing the 60 minutes between 2-3 AM), or when the hour between 1-2 AM is repeated when we "fall back" at the start of Standard Time.  However, it turns out the reason was simpler than that as the US is a signatory of an agreement drawn up by the World Meteorological Organization calling for Universal Coordinated Time to be used as the standard (i.e., not seasonally adjusted).  However, I can't help wondering why local weather records shouldn't reflect the time that its population is conducting its life in.  And it's curious that Standard time, which lasts about four months in much of the US, is the year-round standard.



Standard time


So what does this mean in practice?  Let's say that a rainstorm at 12:30 AM DST on May 10 kept you awake.  Using Standard Time rather than DST, the NWS would report that the rain fell at 11:30 PM on May 9; however, in the world of Daylight Saving Time that you're living in, you may have been watching TV at 11:30 and not in bed being kept awake by the rain. 

(These parallel time frames bring to mind how life was in the US before time zones were established.  Prior to their creation in 1883 there were two types of time.  True Solar Time, based on the sun's actual position over a city, was precise and used by the general public in conducting its day-to-day activities, with every city having its own time (e.g., when it was noon in NYC it was 12:10 PM in Boston, 11:55 AM in Philadelphia, and 11:50 in Washington, DC).  Railway Time, on the other hand, was used by the railroads for scheduling arrivals and departures of its trains and was the impetus for the establishment of time zones, whereby New York, Boston, Philly and DC all had the same time).


Train conductor

Over the course of a year there are instances of daily records hinging on what happens either during the last hour of the day or the first hour of the next day, and the DST-Standard Time situation creates a "fuzzy" reality.  For example, sometimes there are days during DST when it appears a rainfall record has been set, but using Standard time is the official measure, the amount from midnight-1:00 AM DST is credited to 11 PM-midnight, erasing what appeared to be a record amount (however, there can be instances were adding rain to the previous day could result in a record amount for that date).  Regarding temperature records, I can recall a few instances in early October during the past fifteen years when it appeared a new record for the latest date for a low of 70+ was going to be established only to find that, although the temperature dropped below 70° after midnight DST, it was actually credited to the day before, so no record was set. 

Of course, in the grand scheme of things, this isn't a big deal.  Since it's doubtful many people are aware of this idiosyncrasy it's a case of ignorance being bliss.  However, for those of us who are aware of this quirk it's annoying (like a hangnail). 


No To Daylight Savings Time

Here are a number of other thought pieces I've written about weather measurement:

A Case For Reporting Temperature & Precipitation to One or Two Decimals

Can Trace Amounts of Precipitation Add Up to Measurable Amounts?

Why 2011, Not 1983, Should be Crowned as NYC's Wettest Year








Best & Worst Weather Conditions During Mets Home Openers

Mr. met


It seems fitting that the Mets first home opener in 1962, a season in which they lost 120 games, ranks as one of its worst, with cold temperatures, wind and drizzle.  And how perfect was it that the weather was glorious the year they won their first home opener in 1968?  The charts below focus on the best weather conditions in years the Mets won their home openers, while the worst weather is grouped into games lost and games won.  (But does it really matter what the weather is, just as long as the home team wins?)      


 Chart - best and worst mets home openers


To read about weather and outcome of every Mets home opener since 1962, double click here.


Tom seaver



March 2019 Weather Recap - Old Man Winter Makes a Memorable Exit

March green background


March began with a continuation of unseasonably cold temperatures that arrived in the last three days of February.  This extended Arctic outbreak persisted through the first eight days of the month; the average high/low during these days was 36°/25°, nine degrees colder than average.  This included three days in a row with highs of 32° or colder - the longest such stretch in March since 1984.  Besides the cold, an inch or more of snow fell on each of the first four days of the month - the longest streak of its kind.  The 10.4" that fell more than doubled the amount of snow from the previous four months (10.1"), making this the fourth winter of the past five in which March was the snowiest month. 


After March's harsh beginning, temperatures recovered and the rest of the month was two degrees above average, with the coldest reading being 32° (on two days) - and there was no measurable snowfall.  The month's mildest reading was 75° (26 degrees above average, two degrees shy of the record for the date), which occurred on 3/15.  A second day in the 70s came two weeks later (70° on 3/30).  Overall, the month was 0.8 degree below average, making it the sixth March of the past seven that was colder than average (but of the six years, 2019 was closest to average).


75 degrees

Like the temperature, the month's precipitation (3.87") was also slightly below average. The biggest rainfall of the month, and so far this year, occurred on 3/21-22 when 1.43" was measured.  This rainstorm was book-ended by three days before and five days after that had unusually low humidity; the lowest was reported on 3/24 and 3/26 when a few hours in the afternoon had readings of 13% and 14%, respectively. 


The 11-day stretch between Feb. 26 and March 8 saw the most wintry conditions of the winter.  The five-inch snowfall on March 3-4 was the second biggest on record to fall in temperatures that stayed above freezing (the greatest accumulation in above-freezing temperatures happened last year when 5.5" fell the morning of 4/2).


Chart - winter 2018_2019


Of all the Marches since 1950, March 2019 ranks among the ten with the coldest starts.


Chart - coldest starts to march


Finally, although last March was colder than March 2019 (2.4 degrees below average vs. 0.8 below average) it had no days with temperatures colder than 27°, while this March had four such days (its coldest reading was 18°).  However, this March had eight days with highs of 55° or milder while last March had only three.


Other March recaps: