US Heat Island Error

This graph represents the unmodified temperatures summarized by NASA by rural verses urban categories. (The gray lines represent the statistical error for the temperatures, while the red and black lines represent a 5-year running average. Notice that the rural data set (red, top graph) does not show much warming for the past 120 years, while the urban data set (black, lower graph) shows a steep, continuous increase in temperature.  Over the period it increased by 2.5 degrees. The rural data set, on the other hand, shows very little warming in the last half of the twentieth century and has even declined slightly.

When the two data sets are combined (black, upper graph), the entire US data set still shows a significant increase in temperature for the period. This increase in temperature has nothing to do with global warming and everything to do with changes in land use.

NASA's Goddard Institute for Space Studies (GISS), led by Dr. James Hansen, has attempted to remove the heat island effect through a series of equations. The GISS produced a graph that looked  similar to the rural (red) line above, but still showed the late 1990s and 2000s to be the warmest in recorded history.

However, two Canadian scientists discovered an error in the GISS adjustments. The error was small and NASA immediately corrected it. Once corrected, the late 1990s and 2000s were no longer the warmest period in US history. The warmest period was in the 1930s and 40s, with 1934 being the warmest year on record, and 1998 the second warmest. Six of the ten hottest years occurred in the 1920s-50s, while only four have occurred in the 1990s and 2000s.

Scientists are becoming increasingly alarmed over the statistical messaging of US data by NASA's GISS led by Dr. James Hansen. The 2007 error is not the only questionable data that Hansen has released the past few years. Each time Hansen announces that the GISS has discovered a better way to statistically modify actual US ground temperatures, warming becomes even more pronounced and any cooling less pronounced.

For instance,  a comparison of the data compiled by the Historical Climatology Network (USHCN), for Santa Rosa, California show a cooling trend since 1900, while the same data after Hansen's GISS applies their top secret statistical adjustment shows a warming! USHCN measures temperature by taking daily readings from an immobile temperature station. GISS compiles its reports by collecting the USHCN temperature readings and then subjecting them to secret adjustments, allegedly to correct for artificial influences such as land-use changes. Hansen is the scientist who raised the alarm about global warming in 1988 and has been an extremely outspoken proponent of man-caused global warming every since. So much so that he has said that any company or scientist that does not agree with him should be prosecuted for crimes against humanity. More and more scientists are saying that this kind of demagoguery has no place in science. Certainly not in a person who is in charge of the United States temperature data and refuses to let any other scientist review his data adjusting equations. It is so bad, that the US Senate Committee on Public Works notes that Hansen's former NASA Supervisor, Dr. John Theon that Hansen is an embarrassment to NASA. (See YouTube statement by Senator James Inholf (OK) about Hansen).

This concern lead some scientists to actually plot the USHCN data for the entire United States and compare it against the NOAA/GISS adjusted data. The comparison is shocking. The unadjusted data clearly show that the warming in the last quarter of the twentieth century in the US is substantially less than the 1930-1945 period than the GISS adjusted data would have us believe. The GISS adjustments increase the temperatures of the 1975 to 2000 period and reduce the temperatures of the 1930 to 1945 period.

Having said that, it must be clear that some data adjustment is necessary for climate comparisons and trends. These include "adjustments for changing equipment, and adjustments for changing site locations and/or urbanization.  However, all of these adjustments are educated guesses.  Some, like the time of observation adjustment, probably are decent guesses.  Some, like site location adjustments, are terrible as demonstrated at www.surfacestations.org."

Another study by Economist Dr. Ross McKitrick and climatologist Dr. Patrick Michaels, clearly showed that land use still affects published temperature records. compared population growth, economic development, coal consumption and other socio-economic indicators with the temperature trends for the respective areas. If there were no urban heat island effect, there should be no difference between high growth and low growth areas around the world. Instead, they found a large and significantly correlated relationship between population growth (i.e. land use) and temperature as measured by surface stations in the area. On the other hand, there was no correlation when surface satellite temperatures (which is not affected by heat island) was used for the same areas.  When the surface station data was corrected for the heat island effect they found, the warming since 1989 would be reduced by half! (Read article in Journal of Geophysical Research)

While this issue is not yet settled, the controversy is growing. Stay tuned.