An impact analysis of enhanced-greenhouse climate change on the Australian alpine snowpack

This Thesis is concerned with the sensitivity response of the Australian alpine snowpack to the onset of possible enhanced-greenhouse climatic conditions in the 21st Century. The analysis procedure involved the use of both physical and empirical simulation models, and the various caveats associated with each of these components should be assessed when interpreting the results. A statistical downscaling model was constructed, which converted large-scale synoptic data into daily changes in the alpine snowpack. This snow model was calibrated for the Falls Creek site in the Victorian Alps (elevation 1649 m). The model was able to reproduce observed fluctuations in the observed snowpack when it was driven by largescale atmospheric temperature, humidity and airstream inputs. The research methodology used an extensive archive of daily output from the CSIRO 9- level General Circulation Model (GCM). This model incorporated a Mixed Layer Ocean, and operated at an R21 horizontal resolution. A daily-scale validation of a 24- year 1xCO2 control climatology revealed the existence of a number of biases within the simulated atmospheric fields. The most serious of these was a negative bias in tropospheric temperatures of between 2 C and 5 C. These biases were adjusted, and the GCM was used to drive the statistical snow model. The resulting simulation was successfully validated against observed data. The climate change sensitivity evaluation was conducted by applying a 29-year doubled-CO2 data-set from the CSIRO 9-level GCM to the statistical snow model. The resulting simulation showed an extremely high sensitivity response from the model site, with values such as mean snow cover duration and peak seasonal snow depth decreasing by over 90%. This was largely attributed to a particularly strong warming in the driving GCM of around 4.8 C. A range of further sensitivity perturbations were conducted by varying the input temperature fields (in both the GCM and observed atmospheric data-sets) by one degree Celsius increments. The mean snow model response suggested a quasi-exponential decay relationship, with the first degree of warming producing the strongest reduction in snow duration and snowpack depth. For example, mean maximum snow depths decreased by around 40% when the observed atmosphere was increased by 1 C. These changes were caused by a simultaneous decrease in snowfall and a very strong increase in ablation. Some preliminary impact analysis was conducted on various snow-affected sectors. Within the biophysical context, the snowmelt runoff into the Dartmouth Reservoir of northeast Victoria was calculated using a relatively simple terrain interpolation/snowmelt scheme. The seasonal runoff pattern was then perturbed to simulate an environment in which no alpine snowpack existed. The resulting runoff pattern contained an abnormally high mean winter maxima and a depressed spring inflow volume. A socioeconomic analysis was also conducted into the viability of the Australian winter tourism industry under a range of scenario conditions. A statistical regression relationship was delineated between the duration of the snowpack and visitation numbers at various alpine resorts. The analysis suggested that revenue generation and hence commercial feasibility could be threatened by a moderate reduction in the mean size of the Australian snowpack