zulookt.blogg.se

• TRAINING STRESS BALANCE FORECASTING DRIVERS

TRAINING STRESS BALANCE FORECASTING DRIVERS

Further we applied these environmental factors to a species specific empirical model of stomatal conductance for black spruce to compare differences in predicted water regulation response when large-scale (ESM) data are used as drivers versus localized data transformed using this new site-level downscaling technique. Validation over historical decades shows that this technique provides hourly temperature and vapor pressure deficit data accurate to within 0.7%. Large-scale forecast data from the Community Earth System Model were downscaled spatially then temporally based on the cumulative distributions and sub-daily patterns from corresponding observational data at North Mountain (Cape Breton). The technique is based on a combination of probabilistic downscaling methods and control system theory, which together are used to transform large-scale future climate input (air temperature, humidity) to local scale climate parameters important to plant biophysical processes (vapor pressure deficit). A downscaling framework is presented and applied to physiological and climatic data for projecting future climate resilience of one key boreal tree species, black spruce, in Cape Breton Highlands, Nova Scotia.