University of California Water Resources Center

Flows in tidal marshes arc not described well by the classical equations of open channel flow. The geometry of the marsh complicates analysis of the system. Water depths in tidal marshes are usually small, varying from a few centimeters or even dry at low tide to a meter or more at high tide. At low tides, the flow can be channelized and is often essentially one-dimensional. At high tides the system frequently must be described using a more complete two-dimensional approximation. A model is thus required that permits transitions between differing areas of the system. It must accurately represent the changes of the areas inundated during the tidal cycle. It must also be capable of operating to the resolution necessary to characterize the flow regime in the marsh.

Vegetation within the marsh presents a variable resistance to flow particularly in this range of depths. A parameter like Manning's n is insufficient to characterize the variable friction resistance to flow. A new formulation is required which accounts for head losses within the vegetation as well as flow transitions between the channels and vegetated marsh surface. This formulation will be independent of the model used and is relevant whether the model is one-dimensional or two-dimensional. Factors that contribute to head losses in marshes are: turbulent energy losses, bottom friction, wind stress, drag from the erect plants that obstruct the flow, and secondary currents within the vegetation. The relative magnitudes of these effects have not been quantified impervious research. Direct observation of small scale hydrodynamic phenomena is difficult because of the shallow depths, frequently inaccessible field conditions during high tides and the spatial heterogeneity over the marsh surface. Most computation of flow in tidal marshes is carried out using the lumped friction parameter of classical open channel flow. This parameter is a calibration coefficient and has only a limited physical basis.

An additional problem that must be addressed is the influence of tidal marsh inundation on the circulation within the estuary proper. Typical estuary models build the equivalent of sea wall on the estuary perimeter. Depths at these locations typically do not go to zero and the marsh behind these arbitrary boundaries is excluded from the analysis. A model is thus required that can cost effectively and accurately incorporate these areas into the estuarial analysis.

Urban water supply planning has changed greatly in recent decades, and has generally become a much more technically serious endeavor. (Urban water supply has always been a politically serious endeavor, with abundant sources of uncertainty (Lund, 1988a, b).) Yet for all the serious and fine technical work and research on urban water supply engineering and economics, it often seems that such work has not provided a clear unified approach for combining the many technical measures available for water supply system planning and management. This report seeks to provide such a unified analytical approach, addressing the integrated economical use of yield enhancement, water transfer, and demand management measures in a context of risk and uncertainty from many hydrologic and institutional sources.

[Note: see PDF for proper symbols.] In this study water quality in a set of catchments that vary from 6 ha to almost 1500 ha is investigated. Studying catchments across this large range of scales enables us to investigate the scale dependence and fundamental processes controlling catchment biogeochemical export. The Devil Canyon catchment, in the San Bernardino Mountains, California, has some of the highest atmospheric N deposition rates in the world (40-90 kg ha-1 yr-1 at the crest of the catchment). These high rates of deposition have translated into consistently high levels of NO3 - in some streams of the San Bernardino Mountains. However, the streams of the Devil Canyon catchment have widely varying dissolved inorganic nitrogen (DIN) concentrations, variability, and export. These differences are also, to a more limited extent, present for dissolved organic carbon (DOC) but not in other dissolved species (Cl-, SO4 2-, Ca2+ and other weathering products). As catchment size increases DIN and DOC export first increases until catchment area is ~150 ha but then decreases as catchment scale increases beyond that size. Inorganic nitrogen and DOC also share similar temporal variability within the catchments. The reasons for these phenomena appear to be the dominance of flushing of dissolved constituents out of soil at small scales, the groundwater exfiltration of these flushed materials at intermediate scales and the removal of biologically active materials from streamflow through riparian processes at larger scales. While the particular scale effect observed here may not occur over the same range in catchment area in other ecosystems, it is likely that other ecosystems have similar scale dependant processes. Instream removal processes are a particularly relevant process for understanding the loss processes controlling the fate and transport of nutrients derived from agricultural and urban land uses.

Seepage from rivers and natural channels is a serious problem in the Sacramento Valley of California. Impairment of land use occurs along the Sacramento, Feather, Yuba and Bear Rivers as a result of intermittent periods of high ground water or ponded surface water. When channel levels exceed the adjacent ground surface elevation, water moves through and under the confining levels into adjacent lands. If drainage facilities are inadequate, the soil becomes saturated and water often ponds on the surface. The water damages orchards and perennial and annual crops, and prevents working of land, resulting in delays in or prevention of normal planting of annual crops.

The Lisbon metropolitan region has grown rapidly in population since 1970, due largely to the immigration of people from former Portuguese colonies in Africa and from rural areas of the country in pursuit of higher living standards. Much of this population growth was accommodated clustered high-rise apartment blocks (many unpermitted) in the region west of Lisbon, in the municipalities Oeiras, Cascais, Sintra, and Amadora. These developments were largely unplanned, often did not provide for sewage treatment, and lack adequate mass transit or urban amenities such as parks and other open spaces. Moreover, because the main transport axes run east-west, it is difficult for residents of these apartment blocks to go the relatively short distance southward to the coast (e.g., only 10 km from Cacém to the coast).

This region is drained by a set of subparallel streams (each draining about 20-50 km2), flowing roughly north-south through deeply incised valleys to debouch into the Atlantic between Lisbon and Cascais. With rapid urbanization peak runoff has increased, resulting from larger impervious surfaces and sewage from illegal housing settlements. Many reaches have been canalized within concrete walls to increase flood capacity, eliminating physical habitat complexity, and reducing amenity and recreational values. However, the urbanization has occurred mostly on uplands, leaving the bottomlands of the incised stream valleys in many reaches surprisingly unaltered. For decades, these drainages were largely neglected, managed mostly to convey floodwaters, although in some reaches there was strong informal use of the stream corridor and floodplains (such as garden plots). The Water Framework Directive (WFD) adopted by the EU Parliament (2000), has motivated extension and improvement of the regional sewer network to improve water quality. The WFD requires that all water bodies in member states achieve ‘Good Ecological Status’ by 2015, defined in terms of hydromorphological, biological, and physico-chemical quality elements of stream reaches, based on characteristics documented at reference sites.

Located 15 km west of Lisbon, Ribeira da Barcarena-Jardas drains a 35 km2 catchment, whose uppermost reaches are forested, but otherwise alternates between urbanization and remnant agricultural and open-space uses. With improved sewage treatment and water quality, there is strong potential to preserve and restore ecological functions, consistent with goals of ‘good ecological status.’ As illustrated by the successful urban stream project in Cacém, there is tremendous potential for the stream corridor to provide parkland for the dense urban settlements. Through GIS analysis of remotely-sensed data, and field surveys of water quality, habitat structure, riparian vegetation, and fish populations, an interdisciplinary workshop of graduate students from Berkeley and Lisbon analyzed potential opportunities to enhance ecological values and human access along the stream. Our analysis indicated that implementation of stormwater management strategies via relatively unobtrusive retrofits of small open bits of urban land and floodplain within the catchment could mitigate many of the negative hydrologic effects of urbanization. By virtue of its linear nature, the stream corridors could provide pedestrian and bicycle connections from population centers (now under-served by parklands) to cultural features and to coastal beaches and trails. A trail could inspire similar efforts on neighboring, parallel basins that have undergone similar urbanization pressures and face similar challenges in providing underserved urban populations with access to recreation and contact with nature.

Land subsidence caused by the excessive use of groundwater resources has traditionally caused serious and costly damage to the Los Banos-Kettleman City area of California's San Joaquin Valley. Although the arrival of surface water from the Central Valley Project has reduced subsidence in recent decades, the growing instability of surface water supplies has refocused attention on the future of land subsidence in the region. This report develops a three-dimenslonal, numerical simulation model for both groundwater flow and land subsidence. The simulation model is calibrated using observed data from 1972 to 1998. A probable future drought scenario is used to consider the effect on land subsidence of three management alternatives over the next thirty years. Maintaining present practices virtually eliminates unrecoverable land subsidence, but with a growing urban population to the south and concern over the ecological implications of water exportation from the north, it does not appear that the delivery of surface water can be sustained at current levels. The two other proposed management alternatives reduce the dependency on surface water by increasing groundwater withdrawl. Land subsidence is confined to tolerable levels in the more moderate of these proposals, while the more aggressive produces significant long-term subsidence. Finally, an optimization model is formulated to determine maximum groundwater withdrawl from nine water sub-basins without causing irrecoverable subsidence over the forecasted period. The optimization reveals that withdrawl of groundwater supplies can be increased in certain areas in the eastern side of the study area without causing significant subsidence.

San Pablo Creek drains 42 square miles, debouching into the San Pablo Bay in Richmond, California. In 1919, East Bay Municipal Utility District built a dam in the mid-watershed. The Dam rarely releases water, so the reach downstream (lower San Pablo Creek) has a distinct hydrology driven by runoff from the unregulated, lower, 11.2 square-mile drainage area. Perhaps because flooding is infrequent, and because land-use policies and management have not historically considered low-order channels and their riparian habitat, regulating agencies have spent little time collecting baseline information on the creek. This study seeks to gather such baseline information. The specific questions this study addresses are: 1) What are the key ecological and geomorphic transition zones along the Lower San Pablo Creek? 2) What are the geomorphic, hydrologic, and vegetation characteristics in each of these zones? and 3) What are the discharge estimates for cross-sections in each of these zones?

The results of our study indicate that there are five distinct zones along lower San Pablo Creek: the Upper Alluvial Valley, the Lower Alluvial Valley, the Upper Alluvial Fan, the Wildcat-San Pablo Creeks Alluvial Fan, and the Tidal Flats zones. Results from discharge estimates indicate a wide variance of discharge rates between Rantz, Haltiner, and Wannanen-Crippen methods. A high dominance of non-native vegetation and significant incision in the upper cross-sections indicates potential for future restoration efforts.