ClimateChange Assessment Guide.pdf - University of Waterloo

Estimating Future Local Climates4. Estimating Future Local Climates27This chapter discusses the methods available fordeveloping future local climates for use in climatechange impact assessments. Detailed guidance isprovided in Chapter 6. Methods examined includethose that use GCM output to generate changes inaverage variable values on a monthly basis, that is,the change field method, as well as methods fordeveloping synthetic and analogue scenarios andmore sophisticated methods that rely upon the GCMsin conjunction with downscaling methods/models thatprovide high resolution estimates of future climates.4.1 GCM-Based ScenariosAs discussed in the preceding chapter, GCMs are usedto simulate future climates for a family of emissionscenarios that extend to the end of the 21st centuryat coarse spatial scales. Output from the GCMs isavailable in a spatially gridded format on a long-term,continuous basis. The change field method involvescalculating mean monthly changes in future climatefrom the baseline climate, and applying those changes(e.g., +2.5°C, +10% precipitation for January) to existingclimate data. The change field method can be usedin climate change impact assessments related towater balance or water supply investigations; whereinthe spatial and temporal scales of this approach areacceptable and changes in the frequency and magnitudeof extreme events are not of primary concern.For the change field method, the gridded output fromthe GCMs is summarized as monthly (or seasonal, orannual) averages for each selected climate parameter.Usually 1961-1990 is used as the reference (or baseline)climate and the 2020s (2010-2039 or 2011-2040), 2050s(2040-2069 or 2041-2070) and 2080s (2070-2099 or2071-2100) are used for future time periods. Ratios(e.g., precipitation) or differences (e.g., temperature)between the average for the reference climate and theaverages for the future climate horizons are computedon a monthly basis to derive the net change field byparameter. These change field data have been preprocessedand are available at the Canadian ClimateChange Scenarios Network (CCCSN) for many GCMruns (Chapter 6). These changes are applied to a timeseries of archived daily (or other averaging period)observations from climate stations within the study areato develop the future climate conditions. These “new”climate conditions or scenarios are input to hydrologicmodels to assess impacts of the new climate conditions.Examples of studies conducted in Ontario that haveapplied the change field method are presented inChapter 5, Section 5.1.The change field method has been widely adoptedby water managers due to its ease of use. The primaryadvantage, however, is the the availability of numerousGCMs and associated emission scenarios and thesubsequent ability to generate change fields for mostGCM/emission scenario combinations. This allows thewater manager to investigate the full range of predictedchanges, as well as the most frequently predictedchange, and provides the ability to obtain a detailedunderstanding of uncertainty associated with futureclimate selection. Developing an awareness of this rangeof uncertainty is fundamental to understanding thepotential impacts of climate change on water resources.One of the key limitations of the change field methodfor hydrologic impact assessment is that potentialimpacts of climate change on inter-annual or day-to-dayvariability of climate parameters are not represented.The change field method alters the time series averageswhile the variability inherent in the dataset remains thesame. Changes in sequences of wet and dry days arenot altered by this method nor are patterns of intenseprecipitation events. This can lead to an underestimationof future floods, droughts, groundwater recharge andsnow-melt timing (Bates et al., 2008). As such, themethods described in this Guide are not applicable,or are limited in their applicability, to studies in whichshort-duration and/or peak flow characteristics arethe subject of investigation. Furthermore, as changesto sequences of wet and dry days and the patterns ofintense precipitation events as a result of climate changemay also have a significant effect on long-term flowcharacteristics and water budgets, users should considerthis as a source of uncertainty when evaluating results.As climate science is not yet able to accurately quantifythese changes, the significance of this uncertainty is notwell known.In addition, the coarse-scaled GCM output that lacksthe local-scale parameterization and feedback fromlocally significant features (i.e., topography and surfacewater bodies). GCMs do not have the detailed accuracyat the local scale and are instead more representativeof large-scale, average climate characteristics andpotential changes. Further, limited agreement betweenthe various GCMs on changes to future precipitation