UPDATE : The NWS responds about the station issue, see below. – Anthony
Almost 60% of the contiguous USA covered in snow.
While pundits spin attempts at linking snowfall in the Northeast USA to AGW, much like they do in the summer during heat waves, we find that Nature is just taunting them with snow as far south as the Mexican border in Arizona. And there is more to come, in the next week, we may see snow into Florida. During the last week, 652 new snow records were set in the CONUS as seen in the map below:
The record furthest south, in Paradise, AZ, of 6.3 inches snowfall, beat the old 2 inch record going all the way back to 1896. Paradise, AZ is just 40 miles from the Mexican border. You can see all the snow records yourself here.
And according to the NOAA NOHRSC, 57.5% of the CONUS has snow cover.
I also had a look at temperature records this past week, where there were 92 new record low temperatures all the way into Florida, and only 20 new high temperature records set:
One record in particular, at Jal, NM piqued my interest, because it was in the middle of a bunch of record low temperatures. Not only that, it beat the old record high in 1953 by quite a margin, besting it by 7 degrees:
There are no new high temperature records anywhere close to this station, and it stands out like a sore thumb.The nearest official hourly reporting station in Wink, TX just 26 miles away, shows a high of only 62 on Friday February, 22nd according to this data from Weather Underground sourced from NWS:
Another station to the North, Hobbs, NM, 23 miles to the North, another official NOAA airport station, also shows no new record high on that day:
The weather in the area that day was sunny, mild, breezy, and dry:
I had a strong hunch that this station in Jal, NM may have poor siting which contributed to the new record anomalously warm high on Friday, it turns out my hunch was correct.
According the NCDC metadata, the Jal station is a Class A station, meaning it is part of the climate monitoring network. While not part of the USHCN, it does serve as a station used for infill and pairwise comparisons when trying to homogenize the surface temperature record. The NCDC location metadata was a red flag to me: JAL POLICE DEPARTMENT WITHIN AND 2 MILES SE OF POSTOFFICE AT JAL NM
With just a little research, I was able to locate a photo of the station, courtesy of a survey page from New Mexico State University. The photo showed yet another parking lot weather station:
In the photo above (which I have annotated) the Standard Rain Gauge is clearly visible and what looks like the MMTS temperature sensor shelter on a pole is in the distance near the front of the building. Such placements are typical, they try to get over grass where they can trench a cable back to a window or a wall opening to the display in the office.
A further check of metadata revealed the station is located at 32.1103 -103.1872, within the town according to NCDC metadata, and this Google Earth image:
While that lat/lon puts the station in the parking lot, I note that typically most GPS readings in NCDC’s metadata are good to only about 100 feet. And sure enough, right where I suspected it was, was the telltale shadow of the MMTS shelter. Some annotation was added to the Google Earth image to help you visualize what I know from years of experience doing aerial station surveys.
The photo above has an imaging date of 2/7/2011, seen in lower left – winter time, just slightly over a year ago. Click the image to enlarge it for a closer view.
So to summarize:
1. We have a new record high that is anomalously warm, 96 degrees F. No official nearby stations set any comparable high temperatures or new temperature records in the same time frame. It appears all stations experienced similar warm dry breezy weather that day.
2. We have a NOAA temperature sensor a mere 7 feet from the sidewalk and 16 feet from a large brick building (according to the ruler function in Google Earth and the photo from the NM State survey page).
3. We have a massive parking lot beyond that, and a major road just beyond the parking lot, plus a semicircular drive. Essentially the temperature sensor is surrounded with heat sinks.
4. We have a low albedo surface, dry brown grass, under the temperature sensor in February as evident in the Google Earth photo from a year ago, there’s no reason to suspect this year would be any different.
5. The station is located within the UHI bubble of the town.
So given the sunny dry weather with a lack of nearby comparable temperatures or new records, heat sinks all around, the parking lot, the building, the low albedo of dry grass under the sensor, it seems entirely likely to me that this is a false high temperature record.
I have sent a note to NWS in Midland Odessa to have them investigate.
Update: The original photo of Snow on cacti provided by Scrape TV stated on the Scrape TV article it was from 2013. Alert reader J Philip Peterson pointed out the photo they used was from 2007. I’ve updated the photo at the head of the story for accuracy. – Anthony
UPDATE2: 2/25/13 7AM PST Here is what the NWS Midland says in an email to me:
I did find that JALN5 COOP did erroneously report 96 degrees. Unfortunately, the report did make it to a preliminary report OSOMAF. I checked the database use to compile the record to see if the report was corrected, and the database had updated to show the data as missing. This means that the official record will not include the bad report.
The last available RERMAF online is the latest one in our AWIPS system as well. Historically, our site has only issued RER products for the MAF site (also the only site for which daily and monthly climate products are generated), though expansion in the future is possible.
Though the graphic in the article compares observations to the official record and appears to show a new record at Jal, reports gathered from COOP observers in real time should always be considered preliminary. COOP observations are QC’d daily and at the end of the month before they are submitted as final. Not all COOP sites are ideally sited, but the overriding problem with the Jal report appears to be sensor malfunction. The high temperature data has been edited for bad data for several days.
The record should show that a new record has not been set at Jal.
Thanks for the heads-up and seeking clarification on this issue.
Information Technology Officer
NWS Midland, TX