August has about concluded as a dry spell ensues across the region. This is a good opportunity to take stock of the precipitation over the last month, the summer and calendar year to date.
Precipitation is what we are interested in here, so let’s focus on that part of the official ‘climate’ product for Lawrence issued by the NWS daily and published in the Lawrence Journal World:
CLIMATE REPORT NATIONAL WEATHER SERVICE TOPEKA KS 131 AM CDT THU AUG 31 2017 ...THE LAWRENCE KS CLIMATE SUMMARY FOR AUGUST 30 2017... CLIMATE NORMAL PERIOD 1981 TO 2010 WEATHER ITEM OBSERVED NORMAL DEPARTURE LAST VALUE VALUE FROM YEAR NORMAL ............................................... TEMPERATURE (F) YESTERDAY MAXIMUM 83 87 -4 86 MINIMUM 55 62 -7 69 AVERAGE 69 74 -5 78 PRECIPITATION (IN) YESTERDAY 0.00 0.13 -0.13 0.07 MONTH TO DATE 4.82 4.19 0.63 2.58 SINCE JUN 1 14.12 13.83 0.29 8.53 SINCE JAN 1 27.14 27.72 -0.58 23.17 - INDICATES NEGATIVE NUMBERS. R INDICATES RECORD WAS SET OR TIED. MM INDICATES DATA IS MISSING. T INDICATES TRACE AMOUNT.
As at most locations across the country, an identical NWS local ‘climate’ product is reproduced daily by the local newspaper – in our case, the Lawrence Journal World. Lately, however, there is a caveat published in the newspaper below this NWS issued product: “Due to an NWS rain gauge issue, these numbers are inexact.”
Oops! Somebody has already alerted the local newspaper that the rainfall data is ‘inexact.’ Ignoring the issue of what positive value ‘inexact’ rainfall data is (none), lets have a look at how ‘inexact’ the data actually is. This is easily done by comparing the ‘inexact’ data from the Lawrence Airport with the ‘7-Station CoCoRaHS Lawrence average precipitation.’
The CoCoRaHS Network has already been discussed in three previous posts on this blog earlier in the month, starting with my August 6 post ‘Encore Storm Recap.’ Suffice it to say here that this is an ‘NWS-affiliated’ network of individual volunteer rainfall observers that use a standard 4-inch in diameter rain gauge and measure rainfall at roughly the same time daily (7 AM), then immediately disseminate the data via the web, which is then immediately available in various forms to the public on the CoCoRaHS website. This precipitation data is of excellent quality and consistency, it is measured right here in town at 7 locations (not out at the airport) and I keep track of it. So, let’s compare this data to the ‘official’ Lawrence Airport data that the NWS disseminates and the Lawrence Journal World publishes:
7-STN LWC AVG LWC AIRPORT 7-STN MINUS LWC PERCENT
AUGUST 8.48 4.82 3.66 57%
SINCE JUNE 1 20.21 14.12 6.09 70%
SINCE JAN 1 35.47 27.14 8.33 68%
Yes, the ‘official Lawrence precipitation’ figures are, indeed, ‘inexact.’
In August, the airport reported an ‘official’ amount that was 57 percent of the 7-station Lawrence CoCoRaHS average amount for the month. Somewhere from 2 1/2 to around 3 inches of this monthly difference (I estimate 3.01 inches) is a result of the entirely missing airport reading from the major rainfall event of August 5 (see the post on that event for a discussion of this). The rest of the error for the month of August is a result of erroneously low readings at the airport and sometimes even, perhaps, real local differences in rainfall. Unfortunately, it is difficult to know when the differences are ‘real’ and when they are a result of errors in the measurements of the automated airport gauge.
For the three calendar months of summer since June 1st, the official airport reading is 70 percent of the 7-station CoCoRaHS average.
For the calendar year to date, the official airport reading is 68 percent of the CoCoRaHS average.
Is NWS staff aware of the error and missing data? Yes, of course. If so, why do they not issue ‘MM’ for missing data, as their key at the bottom of their disseminated climate product implies they might? That I really cannot answer definitively, nor can probably anyone at the NWS. However, I do offer speculative answers on a previous post discussing this issue in regard to the August 5 rainfall event.
In conclusion, one can only state definitively the following: Due to incomplete and inaccurate data, the NWS rainfall figures on their ‘official daily climate’ reports (prefixed CLI) for cities and town reporting from smaller airports relying on similar automated rain gauges, are of little to no value. Locally, such airport sites that are also NWS daily climate data sites include Lawrence (LWC), St Joseph (STJ) and Johnson County Executive (OJC).
Other similarly equipped airport sites in the area that report daily on regional climate reports (prefixed RTP) include Topeka Forbes Field (FOE), Manhattan Regional Airport (MHK), Emporia Airport (EMP), downtown Kansas City (MKC), Olathe New Century Airport (IXD) and Lees Summit Airport (LXT).
While these sites may provide accurate data a majority of the time, there are missing days from time to time as well as occasional ‘low-balling’ malfunctions. This ultimately results in the data being of no value, since one cannot say definitively when malfunctions might have occurred. Furthermore, the NWS ignores days when the precipitation amount for the day is missing entirely. Rather, the agency treats the missing daily precipitation data as if it never occurred. This further renders as worthless the cumulative monthly precipitation figures on these NWS ‘climate’ products.