92nd American Meteorological Society Annual Meeting (January 22-26, 2012)

Wednesday, 25 January 2012
How HCN-Corrections to COOP-Data Eliminated Coastal-Cooling Trends in 1970-2010 Summer California Maximum Air Temperatures
Hall E (New Orleans Convention Center )
Amanuel T. Ghebreegziabher, San Jose State University, San Jose, CA; and B. Lebassi-Habtezion, J. E. Gonzalez, and R. D. Bornstein

Analyses of California COOP-site monthly-averaged summer Tmax-trends (over the period of 1970-2005) by Lebassi et al. (2009, in J. of Climate) have been extended by: (a) lengthening the period to 2010, (b) trend-comparisons with newly released HCN data, and (c) calculation of trends in annual Tmax-values. HCN data sets are NCDC-homogenized subsets of the “most trusted” COOP sites; they include 12 (of the 52 COOP sites) in the San Francisco Bay Area and four (of 28) in the Southern California Air Basin (SoCAB). COOP data used as HCN1 data were adjusted by NCDC for the following biases: (a) time-of-observations, (b) spatial inhomogeneity, (c) missing values, (d) changes in thermometer type, and (e) urban warming, while HCN2 data do not include the last two corrections. Comparison of the 35- and 40-year COOP monthly-averaged Tmax-trends at the 16 HCN sites showed a high correlation (0.96) between the two periods. It also showed, however, that as the six inland warming-sites (COOP sites that are also HCN sites) of Lebassi et al. are now generally warming a slightly lower rate than five years ago, the seven comparable coastal-cooling sites are thus now generally cooling at a slightly lower rate. Coastal-cooling was shown by Lebassi et al. as a “reverse-reaction” to the regional warming of inland areas, which triggers greater coastal sea breeze activity, and which thus increased cooling onshore flows.

Comparison of HCN1 and COOP 35-year Tmax-trends shows almost no correlation (0.15) between the two periods, as the HCN1-corrections changed six of the seven COOP cooling-sites into HCN1 warming-sites. Only the site with largest original COOP cooling rate also showed HCN1 cooling. Similar comparisons between the COOP and HCN2 sites showed that the HCN2-corrections changed fewer (only four) cooling-sites to warming-sites (and with lower warming-rates); a low correlation (0.44) thus existed between the HCN2 and HCN1 trend-values.

As many climate-change impacts (e.g., brown outs, heat stress, ozone peaks) depend on extreme Tmax-values, and not just averaged-monthly Tmax-values, the SoCAB distribution of the highest COOP Tmax-values anytime (at each of its 28 sites) during the period from 1970-2010 shows three sub-areas, with a boundary-temperature of 34 C (and with the following ranges): (a) cool coastal (27-34 C), (b) cool mountain-tops (28-34 C), and (c) in-between hot-area (34-40 C). The spatial distribution of the trends in these extreme Tmax-values show decreases up to -0.8 C/dec in the coastal cooling areas and increases up to 0.6 C/dec in the inland and mountain warming areas. Note that these trends are larger than the monthly-averaged Tmax-trends (about ±0.3 C/dec) in Lebassi et al.

Supplementary URL: