How good is the NASA GISS global temperature dataset?

Guest essay by Rud Istvan

It is generally accepted that there are two major land temperature record issues: microsite problems, and urban heat island (UHI) effects. Both introduce warming biases.

The SurfaceStations.org project manually inspected and rated 1007 of 1221 USHCN stations (82.5%) using the 2002 Climate Reference Network (CRN) classification scheme (handbook section 2.2.1). The resulting preliminary paper shows a large temperature trend difference (about 0.1C/decade) between acceptably sited stations (CRN 1 or 2) and those with material microsite problems (CRN 3, 4, 5).

That is a real problem, since only 7.9% of USHCN is CRN 1 or 2. The NOAA solution has been to set up USCRN. This is not yet (AFAIK) being used to detect/correct USHCN station microsite issues in either the NCDC or GISS homogenization algorithms.

What about UHI? The NASA GISS website uses Tokyo to explain the issue and its homogenization solution. One could either cool the present to remove UHI or warm the past (inserting artificial UHI for trend comparison purposes). Warming the past is less discordant with the reported present (the UHI correction less noticeable), so preferred by GISS.

In the Surface Stations supplemental materials (available at www.surfacestations.org) only 14 CONUS stations have pristine CRN 1 siting (1.2%). 4 are labeled urban, 3 are suburban, and 7 are rural. Since these 14 have zero microsite issues, they can be used to examine the GISS UHI homogenization. Both the ‘raw’ and the ‘adjusted’ data can be accessed at www.data.giss.nasa.gov/gistemp. Just click on the monthly chart to go to the station selector page, and enter a station name. The following uses [combined location sources] raw v2, and homogenized v3 (since that is all that is now publically available). Only 13 stations proved usable; Corpus Christi v2 raw (urban) has different lat/lon coordinates than v3 homogenized. That could be a mistake, or it might introduce an unfair comparison. Corpus Christie was therefore excluded; the final GISS CRN 1 sample size is N=13.

Is UHI evident in the raw urban stations compared to rural stations (like the GISS Tokyo/Hachijyo example)? Yes. All three urban stations evidence UHI, for example San Antonio TX and Syracuse NY.

But in suburban Laramie WY or Baker OR UHI is not evident in the raw data–just as, for example, there is no UHI in rural Hobart OK or Fairmont CA.

How good was GISS at removing the apparent UHI bias from raw San Antonio and Syracuse? Hard to tell for sure, but it is evident that the past was warmed some to compensate, just as GISS says its homogenization works.

The third pristine urban station, Savannah GA, was homogenized so much its raw UHI warming trend was fully removed. That might make sense given Savannah’s coastal location, moderated by ocean proximity.

GISS should logically leave non-UHI suburban and rural stations relatively untouched. Oops. GISS homogenization cooled the past to add a spurious warming trend to all but one pristine station. For example these two:

In some cases the past was cooled AND the present warmed, as in Laramie WY.

A spurious warming trend was introduced into all three suburban and 6 of 7 rural CRN 1 stations. Only Apalachicola FL emerged from GISS unscathed.

Automated homogenization algorithms like GISS use some form of a regional expectation, comparing a station to ‘neighbors’ to detect/correct ‘outliers’. BUT 92% of US stations have microsite issues. So most neighbors are artificially warm. So the GISS algorithm makes the hash illustrated above. How could it not? And by extension NCDC, BEST, Australian BOM, …