New Mexico Geological Society Annual Spring Meeting — Abstracts

Much of Arizona and New Mexico receive most of their precipitation during the summer months (late June through late August/early September) as a result of a monsoonal flow centered over northwest Mexico. The monsoon is triggered by the reversal of middle and lower tropospheric winds between June and July, which is directly related to the land-ocean temperature/ pressure gradient. During the cooler seasons the winds blow from the landmass (higher pressure) to the ocean (lower pressure); however, during the warmer seasons, the land heats up faster than the ocean, and the winds blow from the ocean (higher pressure) to the landmass (lower pressure).

The Mexican Monson is strikingly similar to the Indian summer monsoon in all od these respects. One aspect of the Indian Monsoon that has been studies for over 100 years is its inverse relationship to the snowpack and snow cover of the Himalayas and Tibetan Plateau. The Indian Monsoon following winters with greater than average snow cover tends have less precipitation than normal and affect a smaller area than normal. The increased soil moisture from melting and runoff in the spring and summer following a winter with above average snowfall prevents the land from heating up as efficiently, thereby suppressing the land-ocean temperature/pressure gradient.

We hypothesize that this is also the case with the Mexican Monsoon, and that summer precipitation over the soutwestern United States might thereby be predictable. As an initial step in examining this hypothesis, we have obtained time series of monthly mean precipitation data from 21 stations in New Mexico, focusing the analysis on July precipitation. Interannual fluctuations of July precipitation from these stations are positively correlated with each other and with the statewide average for New Mexico.

As indices of snow variability we have used Principal Component Analyses of the interannual variability of snow cover in North America (derived from satellite observations) from the winter of 1970/41 through the winter of 1992/93, obtained from the Canadian Climate Centre. Specifically, the Spring and Annual PC's describing variability in the Central/Southern Rocky Mountains were correlated with the July precipitation data from New Mexico stations. A large majority of these correlated coefficients are negative, confirming that July-average precipitation in New Mexico tends to be inversely related to the amount of antecedent snow cover in the Rockies to the north. The correlation coefficients are approximately -0.4, satisfying a one-tailed t-test for statistical significance.