Some of the research activities in which Tony Cahill is involved

Land-Atmosphere Hydrology

Land-atmosphere hydrology deals with the question of how evaporation occurs, and what the effects of evaporation are on the hydrologic cycle and the larger climate cycle.

Some of my recent work in this area has been my involvement in a series of field-scale campaign experiments run by the USDA and NASA. The Southern Great Plains 1999 (SGP99) and the Soil Moisture Experiment 2002 (SMEX02) experiments consist of aircraft-based measurements of regional soil moisture, along with numerous supporting measurements by different researchers. I am one of those different researchers taking supporting measurements. My work consists of measurement of field-scale evaporation, and then trying to tie it (in some manner) to the soil moisture measurements made by the aircraft.

Time of Concentration for Low Slope Areas

Time of concentration can be thought of as the time it takes water from the hydrologically most-distance point in a watershed to reach the outlet. Current methods of estimating time of concentration use the slope of the land surface in the denomenator, so that time of concentration becomes infinite as the slope goes to zero. This contradicts what is seen in reality. Through numerical modeling and validating laboratory experiments, we aim to develop a new predictive model of time of concentration on areas with negligible slope.

Stochastic Hydrology and Wavelets

On a total different note, I have been doing work lately on the uses of wavelets to analyze streamflow time series. Wavelet analysis is like Fourier analysis, only the basis functions are not sines and cosines, but strange functions that often don't have a closed form. One of the most useful things about wavelets is that the wavelet transforms preserves both time and frequency information, unlike the Fourier transform, which preserves only frequency information.

This may all seem fairly abstract, but let's just say that one of the things you can do with the wavelet transform is test whether the variance of a time series is changing, and determine when is it changing. I ran this test on a number of stream flow time series from across the US, and found that stream flow is becoming more variable, probably with respect to the size of individual storms.

Please note that this page is continuously subject to revision and additions. Last updated: 9/29/02