Search

The goal was to build models for predicting ground-level biologically weighted ultraviolet radiation (UV index, shortened to UV here) that would not require substantial execution time in weather and climate models and yet be reasonably accurate. Recent advances in modeling data make this goal possible. UV data computed from Brewer spectrophotometer measurements at Toronto were matched with observed meteorological predictors for 1989-93. Data were stratified into three sets by solar zenith angle 70 [degrees] and time between 1000 and 1400 LST. Stepwise linear regression (SLR) and CART (nonlinear) tree-based regression models were built for UV and N(UV) (ratio: observed UV to clear-sky UV). CART models required fewer predictors to achieve minimum error, and that minimum was lower than SLR. For zenith angle less than 70 [degrees] CART regression models were superior to SLR by 5%-10% error after regression. The CART model had 31% relative error (ratio: estimated mean-squared error after regression to sample variance) and three predictors: total opacity, liquid precipitation, and snow cover. Including five next predictors decreased error only another 1%. For zenith angle 70 [degrees] or greater, SLR could not produce a useful model, whereas CART gave a model with 15% relative error using three predictors. Total opacity is by far the most important predictor throughout. Snow cover enhances UV at the ground by 11%-13% even in cloudy conditions, but its relative influence decreases with zenith angle. For general use at other locations models with as few predictors as possible are desirable. CART models with 34%-35% relative error were built with three predictors: total opacity, zenith angle, and clear-sky UV. Tests were done at 11 stations for several months in 1995. Averaged root-mean-squared discrepancy between predicted and observed UV is reduced about 40% when observed opacity is used for the CART prediction compared to using clear-sky UV. When an 18-h forecast opacity is used the reduction is about 25%. Improvement over clear-sky UV is substantially greater than this on cloudy days. Thus, CART three-predictor models for N(UV) can be used poleward of Toronto in a variety of cloud conditions in analysis or forecast modes. A predictor representing smoke from forest fires was not included. Several cases during the test period showed clear-sky UV was reduced by smoke 30%-50% near to the fires and 20%-30% far downwind.

Classification and regression trees (CART) decision trees analysis of the Vancouver-lower Fraser River valley region, Montreal urban area, and atlantic region gave results consistent with the available data, and the rules for nodes were found to be physically realistic. The regulating role of moisture content of the air on the maximum surface ozone concentration when all other factors are high values is revealed by the analysis.

This paper represents a stage within a larger project to estimate acid ion deposition from cloud impacting on high-elevation forests. Acid ion deposition depends principally on three factors: the liquid water content (LWC). the ion concentration(s) in fog or cloud water, and the efficiency of the deposition process. In the present paper, the objective is to estimate LWC on Roundtop Mountain in southern Quebec from routine meteorological measurements at the Sherbrooke weather station. After describing preliminary efforts, the methodology that was found to work best is presented. This scheme was a hybrid of applications of two statistical nonlinear regression schemes. First, the classification and regression trees (CART) algorithm was applied to predict the occurrence or nonoccurrence of fog at Roundtop. The algorithm produced by this application permitted the elimination of a large proportion of the data records for which fog was very unlikely to occur at Roundtop. The remaining data were then processed by a second application of CART to determine the predictors that are important for estimating LWC at Roundtop. Finally, these same remaining data were processed by the neuro-fuzzy inference systems (NFIS) algorithm to derive the final prediction algorithm. This hybrid method (CART-CART-NFIS) achieved a correlation coefficient of 0.810, with accuracies of 0.962 and 0.664 for the no-fog and fog events, respectively. (Corresponding threat scores were 0.916 and 0.530, respectively.) These measures of skill were significantly better than those obtained from initial estimates or from schemes that used CART alone. Although optical cloud detector and LWC data are necessary for derivation of the fog-occurrence and LWC prediction algorithms, in the end those algorithms are applied to only the predictor data. Fog-occurrence and LWC data are not required, except for verification purposes. The algorithms and list of predictors still need to be tested to determine how widely applicable they are.