*SPECIAL* - A Look at the Discrepancies Between the NWS Official Reporting Sites and Nearby Mesonet Sites

Subjects: Kentucky NWS Official Reporting Sites
                Adjacent Kentucky Mesonet Sites
Time Period: June - September 2014
Topic: Number of 90 degree days

First, let me throw several sets of numbers at you. The data you are about to see is a sampling of the number of this year's 90 degree days for adjacent or nearby Mesonet sites compared to that of the 'official' National Weather Service sites.

Ky Mesonet Caldwell County 12 days
Ky Mesonet Graves County 19 days
Ky Mesonet Marshall County 19 days
NWS Paducah   43 days

Ky Mesonet Oldham County 4 days
Ky Mesonet Shelby County 1 day
NWS Louisville  35 days

Ky Mesonet McCreary County 1 day
Ky Mesonet Casey County 2 days
NWS London  17 days

Ky Mesonet Fayette County 1 day
Ky Mesonet Madison County 1 day
Ky Mesonet Clark County 6 days
Ky Mesonet Nicholas County 12 days
Ky Mesonet Lincoln County 10
NWS Lexington  21 days

Ky Mesonet Barren County 16 days
Ky Mesonet Warren County(1) 27 days
Ky Mesonet Warren County(2) 36 days
NWS Bowling Green  41 days

Ky Mesonet Owsley County 5 days
Ky Mesonet Knott County 4 days
Ky Mesonet Breathitt County 7 days
NWS Jackson  7 days

Ky Mesonet Franklin County 4 days
NWS Frankfort  13 days

Remember, these numbers are for adjacent counties, sometimes within the same county, and nearest Mesonet sites in relation to the 'official' NWS sites. Except for the NWS at Jackson, all of the other 'official' NWS sites show large variations in the number of 90 degree days compared to other reporting stations of the Kentucky Mesonet.

I think you would agree that we need good continuity of data to provide fair representation for the state of Kentucky's climatological record. Of course, this does not have to pertain to the number of 90 degree days only.

Most, if not all, of Kentucky's NWS official reporting stations are located at airports. Now, each site must follow stringent guidelines about siting and exposure of equipment.

For example, the siting of temperature sensors should include locating them at least 100 feet from any extensive concrete or paved areas, or 500 feet from any building or area that might influence readings. Avoid swampy locations where water collects or artificial irrigation areas. Keep grass or vegetation within 100 feet of the site cut to less than 10" in height and provide unobstructed flow of air.

Special attention is to be given to any changes made in the station sensor that could affect data and necessitate the requirement for a temperature comparison routine.

Recently, complaints about the Lexington temperature sensor had technicians busy. I corresponded with one of the engineers and he told me they had conducted numerous tests and that the sensor met quality standards. Local media were invited to have their cameras rolling as to demonstrate the calibration process and show that the equipment was within appropriate quality assurance standards.

Yet, Lexington's sensor has altered things a little bit. But not as much as one thinks. Comparing data from this year to the same time last year (when the old sensor was still in play) Lexington NWS 2014 average high temperatures were 3.05 degrees higher than the nearby Mesonet site versus 2.65 degrees higher last year under the old sensor. But, this is compared to the Mesonet site a few miles away. Keep in mind,  the Lexington NWS recorded 21 days in the 90's. The Mesonet site recorded just 1 during the year 2014. Last year, Lexington NWS saw 17 days in the 90's. The Mesonet site again only recorded 1 day.

As a side thought, I do not understand why 'official' records should be kept at an airport, as this is not climatologically indicative or reflective of the surrounding area, which includes much more grass, trees, and a lot less concrete (attention NWS Louisville @ the Airport).

Well, what about the Mesonet sites? By analyzing the data above, it seems apparent that their data and means of collecting that data follow a different set of standards or guidelines. But, do they?

At least three times a year, technicians visit and perform required tasks to make sure the equipment still meets manufacturer's specifications. Any faulty sensor or equipment is sent back to the laboratory to be either recalibrated or decommissioned.

I wonder about the sites themselves, though. Why is there so much of a temperature variation from site to site, especially those in close proximity to one another?

For example, in Warren County, one site registered 36 days of at least 90 degrees. A few miles away, another site registered 27 days. For me, that's significant. One has to wonder if the site is not right for the equipment or the methods being followed for the equipment's data monitoring is substandard.

If the equipment's sensors fall within the range of quality standards, then there has to be either a site issue or a methodology issue when it comes to the various Mesonet sites across the commonwealth.

I do not know what it's going to take to get someone's attention to make sure that the data collected is fairly representative and contributes to the overall homogeneity reflecting the current affairs of our climate. But when one sees the wide discrepancy between each of the sites I presented to you, would you accept these numbers as legitimate?

And this is just one aspect that I'm looking at, the number of 90 degree days. What about the other measurements that are taken? Wind, precipitation, dewpoint readings? Are these fairly representative when we compare the official data versus the Mesonet sites. Well, I may investigate those as well. But, I think this will do for now.

References:
http://www.nws.noaa.gov/directives/sym/pd01013002curr.pdf
http://www.nws.noaa.gov/wsom/manual/archives/NB118209.HTML#5.3
http://www.kymesonet.org/instrument.html#!instrument
http://www.kymesonet.org/instrument.html#!sitemaint
http://www.kymesonet.org/historical_data.php
http://www.nws.noaa.gov/climate/index.php?wfo=lmk

MS