The work on this California Applications Program effort is managed by:
Randy Hanson, US Geological Survey - San Diego District
Mike Dettinger, US Geological Survey - San Diego District and Climate Research Division, Scripps Inst of Oceanography
 

CAP is working with selected agencies to develop, assess, and incorporate long-lead climate forecasts into California's water resources and hazards management processes. A requirement of water-resource agencies all around the Sierra Nevada is development of seasonal-scale forecasts and strategies for managing the April through July (or September) streamflows in their respective rivers and into their respective reservoirs. Elsewhere in the State, ground-water aquifers are used as important components of water-supply systems. In these settings also, long-lead seasonal forecasts of climate, runoff, and ground-water recharge are key to improved decision-making and managment of the State's (usually overstressed) water resources.

In southern California, CAP members Randy Hanson and Mike Dettinger are carrying out simulations of ground-water/surface-water responses and management. These simulations, for the Santa Clara-Calleguas Basin, Ventura County, California (1950-1995), are forced by 45-yr ensembles of global climate model hindcasts. The Santa Clara-Calleguas region is depicted below.

The USGS developed the ground-water/surface-water model for the United Water Agency of Ventura (depicted below) during the 1990s and it is used to address management issues concerning the impacts of ground-water and surface-water developments and usage in various parts of the basin.

CAP is coordinating uses of this model with United Water and, more recently, Ventura County and the Fox Canyon Ground-water Management Area, in order to design our efforts and experiments to address the most recent management issues. Climate variations influence ground-water and surface-water systems to differing extents. The overall extent of climatic influences in a coupled ground-water/surface-water system can be difficult to guess and depends on how closely the ground-water and surface-water systems are coupled, on which system dominates the year-to-year hydrologic variations in a basin, and on the time scales and avenues by which the climate forcings enter the particular hydrologic systems. Thus the extent to which climate forecasts will be useful in management of ground-water/surface-water systems is uncertain.

Working with the local management agencies, in order to explore the applicability of climate simulations and forecasts in management of the ground-water/surface-water system of this typical West Coast coastal basin, ensembles of hindcast climate simulations from 1950 to 1998, by three different climate-prediction model runs provided by IRI, have been used to force a calibrated model of ground-water/surface-water conditions in the Santa Clara-Calleguas basin of southern California. The image above shows the scale relationship between the climate model and the Santa Clara-Calleguas ground-water/surface-water model. Comparison of the responses simulated with the hindcast fields with historical observations and calibrated simulations of ground-water and surface-water variations in the basin indicate that current climate simulations (and, perhaps, forecasts) reproduce at least some of the longer term forcings to which the ground-water/surface-water systems respond on interannual time scales. The simulations show that the combined climate-model/surface-water/ground-water simulations reproduce the most basic responses of the Santa Clara-Calleguas basin to global climate as forced by the AMIP sea-surface temperatures; that is, the same time scales of variability (QBO, ENSO, and decadal) are traceable from the SST forcings through the simulated Californian precipitation totals through the simulated streamflow and pumpage into the simulated ground-water levels. Furthermore, the magnitude of the mean ground-water responses to the historical range of tropical sea-surface forcings is entirely realistic.

The figure above summarizes how the climate ensembles respond to the tropical sea-surface temperatures that are used to force a climate model (top panel) and how, in turn, the ground-water/surface-water model responds to those climate simulations (bottom panel). The top panel shows that on average (by the regression lines) the ECHAM (climate) model tends to do a good job of returning the observed variation of precipitation over Ventura as a function of imposed tropical forcings (that is, the regression lines lay over each other about as well as anyone could hope). The ECHAM ensemble precipitation members are also observed to have about the same scatter as the observations around the central precipitation tendencies. So it would appear that the ECHAM simulations of precipitation in this case are a reasonable representation of how variable precipitation over Ventura is, both due to ENSO and other "random" variations.

Once the simulated precipitation is used to force the surface-water/ground-water model, the response of the ground-water levels (shown in the bottom panel) can be assessed in much the same way. As with precipitation, the "mean" ground-water level changes in response to various tropical forcings (as indicated by the regression lines) are very similar. Thus the "expected" value of ground-water level response (as a function of tropical forcing) is well simulated by the coupling of climate/ground-water models. The scatter around those expected values are, however, much smaller in the simulation than in the observations. This suggests that somewhere in the process, sources of variability have been neglected or underestimated. The top panel indicates that the climate model is not underrepresenting the precipitation scatter, so that it appears that the current simulations are either underestimating the precipitation effects on the ground-water system (we believe that this could be an underestimation of how much pumpage changes with precipitation), or that other non-precipitation-forced variations, not represented in the model, are missing (like changes in management styles and land uses), or--most likely--that both of these problems are limiting the simulated variability of ground water as a function of tropical forcing.

Initially though, CAP is interested in how well the mean response can be forecasted. In ensemble forecasts of the ground-water responses to forecasted tropical forcings, the results above suggest room for much optimism: the coupled climate/ground-water model system reproduces the historical "mean" ground-water responses remarkably well. The deficit of residual variability around these mean responses in the ground-water ensembles mostly will limit the amount of ensemble ground-water scatter in a forecast and not the mean forecasted ground-water response to a given season's (forecasted) weather. CAP hopes to improve the simulated scatter in order that the uncertainties in such forecasts can be realistically represented and applied in decisionmaking processes.

Forecasts by such methods are thus most immediately limited by the skill of current long-lead precipitation forecasts. The next steps are to resimulate the responses with seasonal, rather than annual, precipitation forcings, and to explore more complex forms of downscaling since the present simulations use precipitation rates directly from the climate models.