Investigating the mesoscale impact of artificial reservoirs on frequency of rain during growing season

Abstract

The specific question that this study pursued is “Have large dams modified the downwind frequency of rainfall in the mesoscale during growing season?” Rigorously quality controlled precipitation data comprising 3055 stations from the Global Historical Climatology Network (GHCN) were analyzed with 92 large dams in the U.S. Using 30 years of atmospheric reanalysis data, the wind rose diagram for each dam was derived from wind data at the 850 mb level. Around 96 (78) GHCN downwind (upwind) precipitation stations were identified that were within 100 km (mesoscale) of dams. The Mediterranean and humid subtropical climates were found to have experienced the highest and statistically significant change in trend in precipitation frequency downwind and within 100 km of dams during the growing season. The warm summer continental climatic region was found to have exhibited the next most modification. Paired analyses were performed as a function of predam and postdam and at upwind and downwind locations. For Mediterranean climates, the stations studied were found to have experienced a generally weak trend in precipitation frequency before the construction of the selected dams and a systematically more impacted trend during the postdam period. However, using precipitation observations alone, the specific role played by irrigation dams could not be distinguished from other types of dams in this study. Analysis of humidity records, however, revealed that dams can increase the moistening of the air mass by about 5%–15% (in terms of vapor pressure) as it passes downwind, while the effect can also be marginal for other dams. In summary, our study reveals that it is easier to establish a physically intuitive connection between large dams and downwind frequency of rain, but it is much more difficult to demonstrate this connection consistently for all the downwind stations in the mesoscale without the use of additional geophysical data (e.g., topography, land use, and land cover patterns) and mesoscale atmospheric modeling.