ID | 1915 |
---|---|
Citation | Water Supply and Demand Working Group. 2010. Okanagan Water Supply and Demand Project: Phase 2 - Appendix N: Development of Climate Data Sets. Prepared for the Okanagan Basin Water Board. |
Organization | OBWB |
URL | http://www.obwb.ca/obwrid/docs/340_Appendix_N_Climate_Data_Sets_Neilsen.pdf |
Abstract/Description or Keywords | This document describes the work done to develop the climate datasets used in the Phase 2 Project. A climate layer consisting of a 500 m by 500 m grid was developed by the University of Lethbridge, Agriculture and Agri-Food Canada and Environment Canada for use in water demand modelling and other applications. Estimates of daily maximum and minimum temperature, as well as daily precipitation, have been derived for the mid-point of each climate grid cell using observed data, for the period 1961 - 2006. For the purposes of this project the period 1996-2006 was examined in more detail. In addition, 12 data sets consisting of daily temperature and precipitation estimates have been generated for each day going back to 1961 and ahead to 2100. These 12 datasets were derived using 6 different General Circulation Models (GCMs), each running twice. These various climate layers have been used in the Phase 2 project for both water demand modelling and water supply modeling. The terrain of the Okanagan Basin is diverse and has a strong influence on climate. Water supply and demand models require climate data inputs that reflect this complexity, but climate stations are few and located mainly in the valley bottom. Consequently, spatial interpolation at a suitable scale is required to fill in gaps in temperature and precipitation data. A number of approaches have been used previously to develop models which incorporate spatial correlation and topographic effects on climate data. These include GIDS (Gradient plus Inverse-Distance-Squared), which weights predictions derived from a multiple regression equation with the inverse distance to nearby climate stations within a specified radius (Nalder and Wein, 1998) and PRISM (Parameterelevation Regressions on Independent Slopes Model) (Daly et al.1994), which takes into account elevation but also topographic facet (aspect). Other approaches include DAYMET (DAilY METeorology) (Thornton et al., 1997), which uses weighted observed climate values within a search radius using a Gaussian filter with a shape parameter and ANUSPLIN (Australian National University SPLINe) (Hutchinson 1995, 1999) which uses a thin plate spline algorithm to fit a smooth surface through data values. Each of these methods has shortcomings in mountainous regions where terrain variations have a substantial effect on local climate. To account for temperature inversions (warm air overlying cold air), two atmospheric layer models have been used. For example, Daly et al. (2003) divided meteorological point data into two sets and used a standard inversion height (based on radiosonde data) and predefined inversion locations were specified. In addition to terrain, large water bodies can also influence climate patterns significantly. Perry and Hollins (2005) used the surface area of water within a 5 km radius of each climate station as a predictive variable for gridding monthly climate surfaces for the United Kingdom. Incorporating the area of surface water around each climate station in the regression-interpolation methodology worked well in all seasons, with the exception of the summer months. In this study, we have created an interactive model for deriving gridded estimates of daily Tmin (minimum temperature), Tmax (maximum temperature), and precipitation for the Okanagan Basin (Figure 1.1). An inverse distance weighting (IDW) interpolation algorithm similar to GIDS was used in conjunction with regional linear and non-linear regression to generate the climate surfaces from weather station data. In addition to latitudinal and elevation influences on temperature and precipitation, temperature inversions are accounted for using a two-layer method and lake effects are modelled by establishing the average long term temperature modification on station Tmax values. climate change, climate data This study builds on past work that investigated the impacts a warmed climate could have on water resources in the Okanagan Basin (Cohen and Kulkarni, 2001; Neilsen et al., 2001; Cohen et al., 2006; Neilsen et al., 2006). The climate surfaces used in these studies were based on the monthly climate grids produced using PRISM. At a spatial scale of 4 and 1 km, the PRISM grids were too coarse to differentiate valley and mid-slope locations (Neilsen et al., 2001; 2006). |
Information Type | report |
Regional Watershed | Okanagan |
Sub-watershed if known | |
Aquifer # | |
Comments | |
Project status | complete |
Contact Name | Denise Neilsen |
Contact Email | [email protected]ᅠ |