2009 temperatures by Jim Hansen

Figure 3. Temperature anomalies in 1998 (left column) and 2005 (right column). Top row is GISS analysis, middle row is HadCRUT analysis, and bottom row is the GISS analysis masked to the same area and resolution as the HadCRUT analysis. [Base period is 1961‐1990.]

Figure 3 shows maps of GISS and HadCRUT 1998 and 2005 temperature anomalies relative to base period 1961‐1990 (the base period used by HadCRUT). The temperature anomalies are at a 5 degree‐by‐5 degree resolution for the GISS data to match that in the HadCRUT analysis. In the lower two maps we display the GISS data masked to the same area and resolution as the HadCRUT analysis. The “masked” GISS data let us quantify the extent to which the difference between the GISS and HadCRUT analyses is due to the data interpolation and extrapolation that occurs in the GISS analysis. The GISS analysis assigns a temperature anomaly to many gridboxes that do not contain measurement data, specifically all gridboxes located within 1200 km of one or more stations that do have defined temperature anomalies.

The rationale for this aspect of the GISS analysis is based on the fact that temperature anomaly patterns tend to be large scale. For example, if it is an unusually cold winter in New York, it is probably unusually cold in Philadelphia too. This fact suggests that it may be better to assign a temperature anomaly based on the nearest stations for a gridbox that contains no observing stations, rather than excluding that gridbox from the global analysis. Tests of this assumption are described in our papers referenced below.

Figure 4. Global surface temperature anomalies relative to 1961‐1990 base period for three cases: HadCRUT, GISS, and GISS anomalies limited to the HadCRUT area. [To obtain consistent time series for the HadCRUT and GISS global means, monthly results were averaged over regions with defined temperature anomalies within four latitude zones (90N‐25N, 25N‐Equator, Equator‐25S, 25S‐90S); the global average then weights these zones by the true area of the full zones, and the annual means are based on those monthly global means.]

Figure 4 shows time series of global temperature for the GISS and HadCRUT analyses, as well as for the GISS analysis masked to the HadCRUT data region. This figure reveals that the differences that have developed between the GISS and HadCRUT global temperatures during the past few decades are due primarily to the extension of the GISS analysis into regions that are excluded from the HadCRUT analysis. The GISS and HadCRUT results are similar during this period, when the analyses are limited to exactly the same area. The GISS analysis also finds 1998 as the warmest year, if analysis is limited to the masked area. The question then becomes: how valid are the extrapolations and interpolation in the GISS analysis? If the temperature anomaly scale is adjusted such that the global mean anomaly is zero, the patterns of warm and cool regions have realistic‐looking meteorological patterns, providing qualitative support for the data extensions. However, we would like a quantitative measure of the uncertainty in our estimate of the global temperature anomaly caused by the fact that the spatial distribution of measurements is incomplete. One way to estimate that uncertainty, or possible error, can be obtained via use of the complete time series of global surface temperature data generated by a global climate model that has been demonstrated to have realistic spatial and temporal variability of surface temperature. We can sample this data set at only the locations where measurement stations exist, use this sub‐sample of data to estimate global temperature change with the GISS analysis method, and compare the result with the “perfect” knowledge of global temperature provided by the data at all gridpoints.




Meteorological Stations




Land‐Ocean Index




Page 2 of 6 | Previous page | Next page