Professional Soldiers ® - View Single Post - An ‘inconvenient result’

vorticity · 08-10-2012, 10:49

Breaking my 'grey-man' lurker status because I actually may have something to contribute to the conversation for once.

Respectfully to QP Pete, Andrew Watts' blog is well-known among climate scientists, and usually can be counted on to 'interpret' data in whatever manner supports his hypothesis that climate change is a.) not influenced by human activity, or b.) not that serious. And using 114 disparate stations over the CONUS to compute a national temperature average, and finding that that average is less than an average computed using a *different* set of disparate stations, isn't all that interesting. He could have easily computed that the USCRN data had a *higher* average temperature, and it also wouldn't mean anything - when you compare stations against each other, you're folding in additional information about location, instrument type and calibration, etc. into the measurement, and that gives you a larger error bar, and differences between the USCRN average and the other average is certainly within that error bar.

What's more useful is an anomaly record for each station - comparing this July's temperature record against all July records measured by that station. Because you're only comparing the station against itself, you don't have to account for differences in sensors, etc. - it's a straight difference comparison, which you can then map out. That's how climate records are measured - not by lumping stations in Wisconsin with stations in Arizona, but by comparing the differences noted by the stations in Wisconsin and Arizona against their own history. You can then take those local anomalies, map them into different regions (national, by state, by county, etc.) and rank them in terms of their anomalies. NCDC has such maps online, and they can be found at:

NCDC Climate Maps - July 2012

There are a lot of questions to resolve about climate and climate change, but the notion that July wasn't the hottest on record for much of the country simply isn't borne out by the data, when making a fair comparison. Hope that's a useful analysis.

08-10-2012, 10:49	#6
vorticity Asset Join Date: Feb 2011 Location: Fort Collins, CO Posts: 15	Breaking my 'grey-man' lurker status because I actually may have something to contribute to the conversation for once. Respectfully to QP Pete, Andrew Watts' blog is well-known among climate scientists, and usually can be counted on to 'interpret' data in whatever manner supports his hypothesis that climate change is a.) not influenced by human activity, or b.) not that serious. And using 114 disparate stations over the CONUS to compute a national temperature average, and finding that that average is less than an average computed using a different set of disparate stations, isn't all that interesting. He could have easily computed that the USCRN data had a higher average temperature, and it also wouldn't mean anything - when you compare stations against each other, you're folding in additional information about location, instrument type and calibration, etc. into the measurement, and that gives you a larger error bar, and differences between the USCRN average and the other average is certainly within that error bar. What's more useful is an anomaly record for each station - comparing this July's temperature record against all July records measured by that station. Because you're only comparing the station against itself, you don't have to account for differences in sensors, etc. - it's a straight difference comparison, which you can then map out. That's how climate records are measured - not by lumping stations in Wisconsin with stations in Arizona, but by comparing the differences noted by the stations in Wisconsin and Arizona against their own history. You can then take those local anomalies, map them into different regions (national, by state, by county, etc.) and rank them in terms of their anomalies. NCDC has such maps online, and they can be found at: NCDC Climate Maps - July 2012 There are a lot of questions to resolve about climate and climate change, but the notion that July wasn't the hottest on record for much of the country simply isn't borne out by the data, when making a fair comparison. Hope that's a useful analysis.
vorticity is offline