Modeling transient pH depressions in coastal streams of British Columbia using neural networks

Transient events in water chemistry in small coastal watersheds, particular ly pH depressions, are largely driven by inputs of precipitation. While the response of each watershed depends upon both the nature of the precipitati on event and the season of the year, how the response changes over time can provide insight into landscape changes. Neural network models for an urban watershed and a rural-suburban watershed were developed in an attempt to d etect changes in system response resulting from changes in the landscape. S eparate models for describing pH depressions for wet season and dry season conditions were developed for a seven year period at each watershed. The ne ural network models allowed separation of the effects of precipitation vari ations and changes in watershed response. The ability to detect trends in p H depression magnitudes was improved by analyzing neural network residuals rather than the raw data. Examination of sensitivity plots of the models in dicated how the neural networks were affected by different inputs. There we re large differences in effects between seasons in the rural-suburban water shed whereas effects in the urban watershed were consistent between seasons . During the study period, the urban watershed showed no change in pH depre ssion response, while the rural-suburban watershed showed a significant inc rease in the magnitude of pH depressions, likely the result of increased ur banization.