Variations in discharge and dissolved organic carbon and nitrogen export from terrestrial basins with changes in climate: A neural network approach

We use a neural network approach to examine relationships between climate and geography on discharge and dissolved organic C (DOG) and N (DON) from 15 river basins in Canada's Atlantic region over a 10-yr period from 1983 to 1992. Model predictions emphasize the importance of the evapotranspiration-precipitation link in establishing basin discharge because even large increases in precipitation can lead to decreasing discharge (Q) when accompanied by higher temperatures. Varying temperature and precipitation inputs to the DOC model revealed that both increases in mean monthly temperature and precipitation will only slightly increase exports from river basins, matching the pattern for e. Surprisingly, lower temperatures and precipitation will, also increase the exports. Increases in temperature with decreases in precipitation will lead to large decreases in export. The DON model predictions were not as accurate as DOC but showed similar tendencies.