Timing, Duration, and Effects of Droughts in the Southern Sierra Nevada and San Joaquin Valley, CA Over the Last 2000 Years

Thursday, 17 December 2015
Poster Hall (Moscone South)
Kenneth D Adams, Desert Research Institute Reno, Reno, NV, United States
The Central Valley of California is one of the most prolific agricultural areas in the U.S., providing about 25 % of the nation’s food. This system is reliant on winter snows in the Sierra Nevada that gradually melt through the spring, but over the last 4 years California has been in the grip of its worst drought of the last 150 years. The question remains, however, how unusual is this drought when compared to previous events over longer time scales?

We used moisture sensitive tree-ring chronologies from the Living Blended Drought Atlas of Cook et al. (2010) to reconstruct annual discharges over the last 2000 years for the Kings, Kaweah, Tule, and Kern rivers in the southern Sierra and routed this discharge into a Tulare Lake water balance model to simulate lake-level fluctuations over this same time period. Although the current drought represents the driest consecutive four year period over the past 2000 years, in terms of discharge volumes, there are multiple periods of more severe, longer term drought represented by extended periods of low lake levels. Significant low-lake periods (< 61 m) include 793-814, 906-933, and 1140-1158, all of which occurred during the Medieval Climate Anomaly. Conversely, lake levels were predominately high during the ensuing Little Ice Age, separated by brief periods of low lake levels. Under natural flow conditions, the 1923-1935 drought would have lowered lake level to about 58 m, which is about 2 m lower than where lake level would have been in the current drought.

Wavelet analyses of the streamflow and lake-level records reveal different periodicities of drought and wet conditions because lake-level is a state variable that changes relatively slowly, depending on inflow, precipitation on the lake, evaporation rate, and the hypsometry of the basin, whereas streamflow is a flux that responds immediately to climate perturbations. The streamflow records have a dominant period of 2-8 yrs but lake-level fluctuations follow longer periods of >32 yrs, primarily prior to 1300. While the 2-8 yr periodicity may reflect ENSO cycles, the causes of the longer periods in the lake-level record remain unknown.