UCD Lake Tahoe Clarity Model: Initial Results Geoff Schladow1, Joaquim Losada, Ted J. Swift, John E. Reuter, Alan D. Jassby Department of Civil & Environmental Engineering, UC Davis Tahoe Research Group, UC Davis      This research had as its goal the quantitative evaluation of the application of coupled hydrodynamic, ecological and optical models, to address the deterministic prediction of water clarity in Lake Tahoe. Prediction of water clarity is somewhat unique, insofar as it represents the integrated and coupled effects of a broad range of individual water quality components. These include the biological components such as phytoplankton, together with the associated cycles of nutrients that are needed to sustain their populations, and abiotic components such as suspended particles that may be introduced by streams, stormwater, atmospheric deposition or sediment resuspension. Changes in clarity induced by any component will feed back on the phytoplankton dynamics, as incident light also affects biological growth.      The choice of model was constrained to a modeling approach that lent itself to the use of the final product as a management tool. Thus model run-time was a crucial issue, a factor that virtually constrained the modeling to be one-dimensional (preserving the vertical direction) in approach. This imposed criterion thus provides a further goal for the research. Namely, to address the question of whether the deterministic prediction of lake clarity can be achieved with the simplifying assumption of one-dimensionality, in a lake that is clearly three-dimensional in nature.      Data from Lake Tahoe - required for model initialization, forcing, calibration and validation - was collated and a semi-automated calibration procedure was developed and used to calibrate the biological and chemical components of the model. The calibrated model was successfully validated using data from a different time period, and is now being used to test hypotheses about the decline of the water quality under given environmental conditions.      Figure 1 shows the good relationship between the observed Secchi depth at the TRG mid-lake station as compared with model prediction for 1999. Actual clarity (as measured by the TRG) varied from a maximum of approximately 34 m to a minimum of approximately 13 m and the model predictions were able to capture this seasonality. Although a complete validation with 1992 was not possible because of the total absence of particle data for that year (and the dominant role that fine particles have on lake clarity) the other measured lake parameters such as chlorophyll and nutrients compared well. Figure 1. UC Davis Tahoe Clarity Model showing close fit between Secchi depth measured in the field and the model prediction for 1999.      Application of the model for potential management purposes suggests some interesting conclusions. These should, however, be taken as preliminary since the model has not as yet been fully calibrated (i.e. do not apply these results for management purposes at this time). For the present work only one-year simulations were conducted, and it is evident that the evolution of changes in water clarity will occur over many years. A comparison of their effects over the long term are presently being considered, with results expected soon.      First, it seems that the fine inorganic particles (mineral or soil origin) are exerting a major influence on water clarity at Lake Tahoe. A reduction in the lakeÕs mineral particle inventory by approximately 50% could return the lake to the clarity levels experienced 30 years ago (Figure 2). At the same time, continued build up of mineral particles will further reduce clarity. Figure 2. Predicted effect on lake clarity if the current number of soil mineral particles in Lake Tahoe were increased by 50% and decreased by 50%. Model is sensitive to changes in mineral particles.      Figure 3 shows that fine particles (clays) in Lake Tahoe need to be lowered by approximately 50% to achieve the TRPA Threshold. A 10% reduction in clays is not sufficient and even if all the coarse particles (silts and fine sand) were removed, there would be a minimal impact on clarity.      The model has shown its value as a tool to examine the driving forces of Lake Tahoe, both in a physical and ecological sense. The model structure allows for testing new hypothesis. For example, the role of Mysis and its interaction in the nutrient budget of the lake, acting as a biological pump of nutrients, has been treated quite simply in the present model. However, a number of different hypotheses could readily be tested. This would need to be done in close cooperation between modelers and field researchers. Figure 3. Predicted change in Lake Tahoe clarity with reduction in clays, silts and fine sand.      Identification of data gaps has been highlighted as result of the modeling conducted to date. As has been demonstrated, the ability to calibrate and validate the model is mainly limited by the poor spatial and temporal resolution of the data. In a broader context, it should be recognized that Lake Tahoe has a relatively high degree of monitoring, and that even greater data issues would exist for most other lakes. One of the most important findings of the model has been to identify the role of the inorganic and organic particles as a key factors in determining the clarity of Lake Tahoe, and the critical importance of determining the relative fraction of inorganic and organic sediments.      Of future research that should be undertaken, there are a number of areas that stand out. The first is the construction of representative long-term synthetic data sets. Long-term simulations require the definition and generation of sets of synthetic data of meteorological variables, stream inputs and outflows. Thus, additional research is needed to define adequate criteria that reflect the range of variability usually found in the lake and its watershed. A range of time scales will need to be captured in this data set, including El Nino effects, drought cycles, climate change and purely random variability.      Secondly, particles interactions are of key concern, because of the dominance that the particles appear to have on light attenuation. Areas that need further work include more data on the particle size distribution and its composition. An equally important area that has not been addressed in the present work are mechanisms of particle aggregation and removal from the water column. While this could be neglected for the one-year simulations undertaken here, long-term simulations are not possible without considering it fully.      For the purposes of producing an operating model in a reasonable time, relatively standard formulations were utilized for primary production and nutrient uptake. While these may indeed be satisfactory, it would be wise to consider some of the newer approaches that are now appearing in the literature. A related area to this is the better definition of the role of zooplankton and Mysis grazing on nutrient dynamics.