11.07.2015 Views

ClimateChange Assessment Guide.pdf - University of Waterloo

ClimateChange Assessment Guide.pdf - University of Waterloo

ClimateChange Assessment Guide.pdf - University of Waterloo

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Appendix E E-47Subwatershed 19 Case Study The Subwatershed 19 climate change impact assessment compared streamflow and groundwater discharge underfuture climates and under the current climate. The assessment requires recognizing uncertainties associated with thestreamflow model, the groundwater flow model, generation <strong>of</strong> future climates, the application <strong>of</strong> the future climates tothe two models, and the statistical representation <strong>of</strong> the results. These uncertainties are discussed below.As all models require the use <strong>of</strong> certain assumptions to simplify the hydrologic and hydrogeologic systems, thesimulated results contain uncertainties. In the Subwatershed 19 HSP-F model, there are uncertainties associatedwith the outflow <strong>of</strong> Island Lake as no continuous long-term records exist. In addition, there are missing data instreamflow datasets, particularly in winter and spring months due to freezing conditions. This adds uncertainty;however it is small as the source <strong>of</strong> missing data likely amounts to less than 1% <strong>of</strong> total measure flow volumes(AquaResource, 2008).In the MODFLOW model, the distribution <strong>of</strong> data points and the poor quality <strong>of</strong> some data (e.g., geological descriptorsin water well records) meant that a number <strong>of</strong> simplifying assumptions needed to be made regarding the geology orthe hydrostratigraphy <strong>of</strong> the system. In an area such as the Orangeville Moraine, the stratigraphy is highly variable,and as the number <strong>of</strong> boreholes used to characterize the geology <strong>of</strong> the area is limited there is a level <strong>of</strong> uncertaintyassociated with the properties applied in the model and whether they are representative <strong>of</strong> the real world conditions.Despite the limitations noted above, the groundwater and surface water flow models were well calibrated andsimulated reasonable estimates <strong>of</strong> water budget parameters, streamflow, groundwater recharge and groundwaterdischarge.There is uncertainty associated with which future climate Subwatershed 19 will experience in the 2050s, andsubsequently with the magnitude <strong>of</strong> hydrologic impacts. To better quantify the range <strong>of</strong> uncertainty associated withmultiple future climates, a number <strong>of</strong> GCM scenarios were investigated. The range <strong>of</strong> future climates investigatedspanned the complete range <strong>of</strong> predictions, from significant warming to slight warming, from wet to dry. Investigatingmultiple future climates allowed the upper and lower bounds <strong>of</strong> impacts to be quantified, as well as the centraltendency. Approaching uncertainty in this manner allows the investigator to determine the impact most <strong>of</strong>tenpredicted to occur in the future, while identifying the extremes.There are significant uncertainties associated with the development <strong>of</strong> local climates from output generated from GCMscenarios. These uncertainties include, but are not limited to: lack <strong>of</strong> local scale features (e.g., Niagara Escarpment,Great Lakes) which affect climate; the inability to replicate sub-scale convective thunderstorm events; shifts inprecipitation intensity being poorly understood; and shifts in drought frequency and duration being poorly understood.In an attempt to reduce these uncertainties, several downscaling techniques were investigated as part <strong>of</strong> this study.In many cases, limitations with the GCM output caused the downscaled climates to be un-realistic, with the mostcommon issue being the drizzle effect. The drizzle effect is related to GCMs predicting small precipitation amountsalmost every day, while still predicting reasonable annual totals. The drizzle effect was subsequently passed on tooutput generated by the weather generator and the regional climate model, and resulted in an implausible futureclimate (precipitation occurred every day). As a result, these downscaling methodologies were not further considered.The presentation <strong>of</strong> results adds a level <strong>of</strong> subjectivity and uncertainty to the climate change impact assessment.Some forms <strong>of</strong> presentation are more subjective than others, such as the histograms, where the ranges <strong>of</strong> each blockare selected by the analyst (e.g., 0-10% vs. 0-5%). Presenting results in the form <strong>of</strong> the box-and-whisker diagrams, isa more objective approach and allows the reader to visualize the inherent uncertainties and ranges <strong>of</strong> possibleimpacts associated with the climate change impact assessment.Despite the uncertainties, the Subwatershed 19 climate change impact assessment is able to provide insight into therange <strong>of</strong> possible hydrologic effects related to climate change.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!