Search

Climate change, in combination with reduced external nutrient loading are important drivers of phytoplankton changes in lakes undergoing re-oligotrophication. Understanding and forecasting future changes in primary production in response to local and global forcing have become a major challenge for developing sustainable lake management. Primary production (PP) in lakes is driven by eco-physiological processes and physical conditions. Thus, for quantifying PP rates, coupled physical-biogeochemical models prove to be useful tools. The objective of this study is to assess how well different coupled models reproduce observed temperature and PP profiles measured at a fixed station (SHL2) in the middle of Lake Geneva. Here we coupled the 1D General Ocean Turbulence Model (GOTM) with various biogeochemical models through the framework of aquatic biogeochemical models (FABM). We performed the first set of simulations with only GOTM for studying the sensitivity of the physical conditions to subtle changes in meteorological variables. Performance of GOTM appears to improve when the heat fluxes due to precipitation and evaporations are represented. Also, we found the model results to be extremely sensitive to SWR and wind intensity. In the next step, GOTM is coupled with three biogeochemical models of varying biological complexity. Based on the comparisons between the performances of the biological models, we aim to gain insight into the optimal complexity of the biogeochemical models to be used. A data assimilation method is used to minimise misfits and estimate optimised parameter values and associated uncertainties for the biological parameters related to phytoplankton growth. The preliminary results show that the complex models, that resolve more physio-ecological processes than a simple NPZD model, perform better in terms of reproducing observations and in yielding lower estimates of parameter uncertainties. The next step is to cross-validate the model solutions with data from a field station in Lake Lugano (independent data, not used for optimisations). The objective behind this cross-validation exercise is to test the robustness of model solutions which is one of the metrics to assess skills of a model. [ABSTRACT FROM AUTHOR]

Measuring primary production (PP) is of major importance to evaluate how lakes are recovering from eutrophication. The classical sampling method involves radiolabelled carbon incubated in bottles at selected depth. Yet the logistic associated with this method prevents scientists from sampling the high frequency dynamics of PP. Alternative methods are based on oxygen sensors and infer PP from diel oxygen variation. To explore the interplay between biogeochemical and physical processes, and in particular PP in large lakes, we have deployed in October 2018 a sophisticated autonomous profiler mounted with optical sensors to measure inherent water optical properties in Lake Geneva. They record backscattering and fluorescence at specific wavelengths as proxies for the composition of suspended particles (phytoplankton and heterotrophic bacteria biomasses), light absorption, attenuation, and backscattering across the visible spectrum, and dissolved oxygen and temperature. The profiler records all parameters every three hours over the top 50 m of the water column with a 10 cm resolution, and provides unprecedented detailed information on temporal and vertical variability of biotic and abiotic processes. Preliminary results show that primary producers' signature on optical properties is large at the sub-daily scale despite low primary production rates in winter. The high vertical resolution of temperature and oxygen profiles allows us to quantify heat and air-water gas fluxes. This in-situ data will be combined with three-dimensional hydrodynamic modelling of Lake Geneva (www.meteolakes.ch), and remotely sensed optical properties to upscale PP estimates from local to basin-scale. [ABSTRACT FROM AUTHOR]