Estimating oceanic precipitation during the 1930s “Dustbowl” and 1950s US droughts

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 19 January 2010: 2:30 PM
B216 (GWCC)
Mathew Barlow, Univ. of Massachusetts, Lowell, MA

Model simulations have suggested that both the 1930s and 1950s US droughts may have been largely forced by tropical Sea Surface Temperatures (SSTs). However, our ability to assess these model results and better understand the drought forcing is limited by our poor knowledge of oceanic precipitation anomalies during these periods. Here we reconstruct and analyze the oceanic precipitation during these two droughts.

It is possible to estimate oceanic precipitation by using global patterns obtained from satellite-era products combined with time series determined by the variability of the land precipitation during the pre-satellite era. This approach has recently been used to provide estimates back to 1949 and back to 1900. It is also possible to apply this approach using time series determined by other datasets, such as land surface temperatures, SSTs, sea level pressure, etc. We have used a multi-variate EOF to relate patterns of the base variable (e.g., SSTs) with patterns of oceanic precipitation for the 1979-2002 period, and then a least-squares fit to the historical variability of the base variable to reconstruct the oceanic precipitation anomalies for 1900-1979. Comparison of the different reconstructions are used to provide an estimate of uncertainty.

For the 1950s drought, there is good agreement on the main areas of Pacific tropical convection anomalies among the different reconstructions, although there is some disagreement in the subtropical western Pacific. For the 1930s “Dustbowl” drought, there is general agreement on a swath of near-equatorial negative anomalies across the Pacific, and an apparent shift in the South Pacific Convergence Zone, but there is considerable disagreement again in the subtropical western Pacific. The temperature-based reconstructions appear to provide additional, perhaps complementary, information in that region. We are currently investigating whether island stations or estimates of low-level wind can be used to better constrain the signal in the western Pacific.