American Society of Civil Engineers

Penman-Monteith Estimates of Reservoir Evaporation

by Marvin E. Jensen, Hon.M.ASCE, (Consultant, Fort Collins, CO, E-mail:, Avry Dotan, (Principal, AD Consultants. Moraga, CA. E-mail:, and Roland Sanford, (General Manager, Mendocino County Water Agency, Ukiah, CA, E-mail:
Section: Irrigation Adaptation to Changing Water Supplies, pp. 1-24, (doi:

     Access full text
     Purchase Subscription
     Permissions for Reuse  

Document type: Conference Proceeding Paper
Part of: Impacts of Global Climate Change
Abstract: Weather and Lake Berryessa (LB) temperature profile data were collected from May 2003 through September 2004 to enable estimating lake evaporation using the Penman-Monteith equation. Current evaporation estimates from LB by the U.S. Bureau of Reclamation (USBR) are based on Class A pan evaporation and the original pan coefficients developed in the 1960s. Since then, the location of the pan has been moved several times and measured pan evaporation measured at the current or old site is much lower than that measured at a new site that is fully exposed to solar radiation. USBR calculated inflow to LB is based on measured lake elevation, estimated evaporation, and measured releases from the reservoir. The estimated lake evaporation based on the pan data and original pan coefficients are too low. Because of the method of calculating lake inflow, calculated daily negative lake inflows commonly occur from mid-July until fall-winter rains begin. Estimates of evaporation from the lake were made using the Penman-Monteith (P-M) equation, estimates of daily change in heat in the water, and advection of heat energy into or out of the lake. Estimates of P-M evaporation were reduced by a factor of 0.95 because a small part of the lake surface area is shaded part of the day, and weather data were measured at a location near the lake instead of over the lake. The P-M evaporation estimates and estimates using pan evaporation at the new site with Lake Elsinore monthly pan coefficients indicated that USBR evaporation estimates are about 20% too low mainly because of pan site conditions. A computer model was developed to process, store, and calculate estimated daily or monthly evaporation from the lake.

ASCE Subject Headings:
Data collection
Hydrologic data