Publications Library

Typically, sub-seasonal to intra-annual climate forecasts are based on ensemble mean (EM) predictions. The EM prediction provides only a part of the information available from the ensemble forecast. Here we test the null hypothesis that the observations are randomly distributed about the EM predictions using a new metric that quantifies the distance between the EM predictions from the National Centers for Environmental Prediction (NCEP) Climate Forecast System version 2 (CFSv2) and the observations represented by CFSv2 Reanalysis. The null hypothesis cannot be rejected in this study. Hence, we argue that the higher order statistics such as ensemble standard deviation are also needed to describe the forecast. We also show that removal of systematic errors that are a function of the forecast initialization month and lead time is a necessary pre-processing step. Finally, we show that CFSv2 provides useful ensemble climate forecasts from 0 to 9 month lead time in several regions.

Climate change detection and attribution studies rely on historical simulations using specified combinations of forcings to quantify the contributions from greenhouse gases and other forcings to observed climate change. In the last CMIP5 exercise, in addition to the so-called all-forcings simulations, which are driven with a combination of anthropogenic and natural forcings, natural forcings–only and greenhouse gas–only simulations were prioritized among other possible experiments. This study addresses the question of optimally designing this set of experiments to estimate the recent greenhouse gas–induced warming, which is highly relevant to the problem of constraining estimates of the transient climate response. Based on Monte Carlo simulations and considering experimental designs with a fixed budget for the number of simulations that modeling centers can perform, the most accurate estimate of historical greenhouse gas–induced warming is obtained with a design using a combination of all-forcings, natural forcings–only, and aerosol forcing–only simulations. An investigation of optimal ensemble sizes, given the constraint on the total number of simulations, indicates that allocating larger ensemble sizes to weaker forcings, such as natural-only, is optimal

Recent improvements in forecast skill of the climate system by dynamical climate models could lead to improvements in seasonal streamflow predictions. This study evaluates the hydrologic prediction skill of a dynamical climate model-driven hydrologic prediction system (CM-HPS), based on an ensemble of statistically-downscaled outputs from the Canadian Seasonal to Interannual Prediction System (CanSIPS). For comparison, historical/future climate traces-driven Ensemble Streamflow Prediction (ESP) was employed. The Variable Infiltration Capacity (VIC) hydrologic model set-up for the Fraser River basin, British Columbia, Canada was used as a test-bed for the two systems. In both cases, results revealed limited precipitation prediction skill. For streamflow prediction, the ESP approach has very limited or no correlation skill beyond the months influenced by initial hydrologic conditions, while the CM-HPS has moderately better correlation skill, attributable to the enhanced temperature prediction skill that results from CanSIPS’s ability to predict El Niño Southern Oscillation (ENSO) and its teleconnections. The root mean square error, bias and categorical skills for the two methods are mostly similar. Hydrologic modelling uncertainty also affects the prediction skill, and in some cases prediction skill is constrained by hydrologic model skill. Overall, the CM-HPS shows potential for seasonal streamflow prediction, and further enhancements in climate models could potentially to lead to more skilful hydrologic predictions.

Infrastructure such as dams and reservoirs are critical water-supply features in several regions of the world. However, ongoing population growth, increased demand and climate variability/change necessitate the better understanding of these systems, particularly in terms of their long-term trends. The Sooke Reservoir (SR) of British Columbia, Canada is one such reservoir that currently supplies water to ~300,000 people, and is subject to considerable inter and intra-annual climatic variations. The main objectives of this study are to better understand the characteristics of the SR through an in-depth assessment of the contemporary water balance when the basin was intensively monitored (1996–2005), to use standardized runoff to select the best timescale to compute the Standard Precipitation (SPI) and Standard Precipitation Evaporation Indices (SPEI) to estimate trends in water availability over 1919 to 2005. Estimates of runoff and evaporation were validated by comparing simulated change in storage, computed by adding inputs and subtracting outputs from the known water levels by month, to observed change in storage. Water balance closure was within ±11% of the monthly change in storage on average when excluding months with spill pre-2002. The highest evaporation, dry season (1998) and lowest precipitation, wet season (2000/2001) from the intensively monitored period were used to construct a worst-case scenario to determine the resilience of the SR to drought. Under such conditions, the SR could support Greater Victoria until the start of the third wet season. The SPEI and SPI computed on a three-month timescale had the highest correlation with the standardized runoff, R2 equaled 0.93 and 0.90, respectively. A trend toward drier conditions was shown by SPEI over 1919 to 2005, while moistening over the same period was shown by SPI, although trends were small in magnitude. This study contributes a validated application of SPI and SPEI, giving more credit to their trends and estimated changes in drought.

To identify agricultural water requirements and the potential for water conservation, we have developed a GIS based model to estimate water demand based on crop type, seasonal crop development, and spatial distribution. Multiple data layers characterize topography, surface and subsurface hydrology, crop distribution, irrigation management, soils, and a range of geographic and political boundaries. The model is driven 500×500 m gridded climate data for the years 1960-2006 and future climate data sets downscaled from GCM output (12 scenarios) from 1961-2100. These climate surfaces have been used to provide calculations of Penman Monteith reference evapotranspiration and a range of agroclimatic indices for each climate grid cell, in addition to crop and terrain based irrigation water demand. Sample scenarios of changing climate, expanded agriculture, and urbanization demonstrate the usefulness of this type of modeling for regional water planning.

While researchers have identified teleconnections between El Niño-Southern Oscillation (ENSO) and extended winter precipitation extremes in North America using generalized extreme value (GEV) models, the regional form of the statistical relationship remains an open question. Past work has shown that relatively warm ENSO conditions may be needed to trigger a nonlinear response over North America. However, studies that stratify winters into La Niña, neutral, and El Niño phases have found that precipitation extremes in neutral/La Niña winters respond differently than in El Niño winters, whereas studies that stratify ENSO data into cold/warm conditions have not found evidence for a coherent nonlinear response. Data and methodological differences have made direct comparison between results difficult. In this study, evidence for a nonlinear association between ENSO and precipitation extremes is reassessed by fitting stationary and linear/nonlinear GEV regression models, with the Niño3.4 index as a covariate, to 1-, 5-, and 10-day extended winter precipitation maxima. Two types of nonlinear model – both one-sided GEV regressions where linear relationships are allowed to differ above/below a Niño3.4 index threshold – are considered. In the first, following past work, the breakpoint is fixed at zero, that is stratifying data into warm/cold conditions. In the second, to see whether there is evidence for a differential neutral/La Niña and El Niño response, the breakpoint is allowed to vary freely. Due to the non-nested nature of the set of models, the Akaike Information Criterion is used to assess the relative support for each model. Depending on accumulation time-scale, the strength of evidence favours nonlinear models at 28%–30% of stations when the Niño3.4 breakpoint is free to vary, versus just 3% with a fixed breakpoint. Optimum Niño3.4 breakpoints are positive (> +0.4°C) in the majority of the nonlinear models, confirming that ENSO/precipitation relationships differs between La Niña/neutral and El Niño winters.

Bolivia's forests contribute to the global carbon and water cycle, as well as to global biodiversity. The survival of these forests may be at risk due to climate change. To explore the associated mechanisms and uncertainties, a regionally adapted dynamic vegetation model was implemented for the Bolivian case, and forced with two contrasting climate change projections. Changes in carbon stocks and fluxes were evaluated, factoring out the individual contributions of atmospheric carbon dioxide ([CO2]), temperature, and precipitation. Impacts ranged from a strong increase to a severe loss of vegetation carbon (cv), depending on differences in climate projections, as well as the physiological response to rising [CO2]. The loss of cv simulated for an extremely dry projection was primarily driven by a reduction in gross primary productivity, and secondarily by enhanced emissions from fires and autotrophic respiration. In the wet forest, less precipitation and higher temperatures equally reduced cv, while in the dry forest, the impact of precipitation was dominating. The temperature-related reduction of cv was mainly due to a decrease in photosynthesis and only to lesser extent because of more autotrophic respiration and less stomatal conductance as a response to an increasing atmospheric evaporative demand. Under an extremely dry projection, tropical dry forests were simulated to virtually disappear, regardless of the potential fertilizing effect of rising [CO2]. This suggests a higher risk for forest loss along the drier southern fringe of the Amazon if annual precipitation will decrease substantially.

Logistical constraints can limit the number of global climate model (GCM) simulations considered in a climate change impact assessment. When dealing with annual or seasonal variables, one can visualize and manually select GCM scenarios to cover as much of the ensemble’s range of changes as possible. Most environmental systems are sensitive to climate conditions (e.g., extremes) that cannot be described by a small number of variables. Instead, algorithms like k-means clustering have been used to select representative ensemble members. Clustering algorithms are, however, biased toward high-density regions of climate variable space and tend to select scenarios that describe the central tendency rather than the full spread of an ensemble. Also, scenarios selected via clustering may not be ordered: that is, scenarios in the five-cluster solution may not appear in the six-cluster solution, which makes recommending a consistent set of scenarios to researchers with different needs difficult. Alternatively, an automated procedure based on a cluster initialization algorithm is proposed and applied to changes in 27 climate extremes indices between 1986–2005 and 2081–2100 from a large ensemble of phase 5 of the Coupled Model Intercomparison Project (CMIP5) simulations. Selections by the method are ordered and are designed to span the overall range of the ensemble. The number of scenarios required to account for changes spanned by at least 90% of the CMIP5 ensemble members is reported for 21 regions of the globe and compared with k-means clustering. On average, the proposed method requires 40% fewer scenarios to meet this threshold than k-means clustering does.

Rainfall extreme value estimates in Canada have historically been based on fitting the Gumbel distribution to annual maxima at individual sites by the method of moments (MOM). Studies have, however, shown that regional frequency analyses (RFA) may perform better than at-site methods. Also, the Fréchet rather than Gumbel form of the generalized extreme value (GEV) distribution may better describe the distribution of annual extremes. In this study, at-site Gumbel MOM and GEV extreme value analyses based on L-moment, maximum likelihood (MLE), and generalized maximum likelihood (GML) estimators are compared against RFA with L-moment methods at stations in southern British Columbia, Canada via cross-validation and Monte Carlo simulations. While GEV shape parameter estimates are predominately negative, qualitatively showing weak evidence for the Fréchet form of the GEV distribution, field significant differences from the Gumbel distribution in the region are not found. Regional frequency analysis leads to substantial reductions in error relative to at-site methods, especially for the GEV distribution and small samples. While Gumbel estimators exhibit lower variance than GEV estimators, they are also more biased, underestimating 100 year return levels. Of the at-site GEV estimators, GML tended to perform better than the L-moment estimator, in some cases nearing performance of RFA. Maximum likelihood performed worst, especially for small samples sizes.