Since the late 19th century, significant changes have occurred in the San Joaquin Valley of Central California, most notably the conversion of over 3 million acres of what was essentially desert into productive farmland. The conversion was possible through extensive irrigation projects by the Federal, State, Local and private entities. The question addressed here is whether such considerable change in the surface characteristics might have an effect on the long term climate record.

To answer this question, we gathered all available data for the region and even digitized many years ourselves. Over 1600 pages of metadata were manually studied and information entered which indicated potential breaks in every station record. At each potential change, a breakpoint was inserted. Thus a single station in effect would be represented by several disjoint segments, or mathematically into several different stations. The stations were divided into two strata, the San Joaquin Valley (<130 m elevation, 18 stations) and the Sierra Nevada Mountains (>130 m, 23 stations). Essentially this separates stations in the irrigated valley from the foothills and mountains of the Sierras. There were 112 Valley segments and 137 Mountain segments. Because of few opportunities for overlapping calculations in the earliest years, the time series will begin in 1910.

A methodology was developed using mathematical graph theory to combine these segments into two regional time series (Valley and Sierra) for TMax and TMin for each of the four seasons. Several tests were performed to test the robustness of the results, especially regarding the 94-year trends.

The results show no consistently significant trends except for Valley TMin which warmed at highly significant rates. The differential warming in spring and summer between Valley and Sierra stations was > 0.40 °C/decade, or about 4 °C over the time period studied - a remarkable temperature change. These are the seasons of largest irrigation deliveries. Overall, annual mean temperature trends were +0.07 and -0.02 °C/decade for Valley and Mountain respectively.

The working hypothesis is that the massive growth in irrigated agriculture during the 20th century changed what was essentially a high-albedo desert into a darker, moister vegetated plain. This altered the heat flux characteristics, allowing for more heat absorption during the cloudless days which would then be released during the night, keeping the Valley warmer than under natural conditions. There were slight decreases in Valley TMax, likely related to evaporative cooling during the day.