AI-based runoff simulation based on remote sensing observations: A case study of two river basins in the United States and Canada

Data-driven techniques are used extensively for hydrologic time-series prediction. We created various data-driven models (DDMs) based on machine learning: long short-term memory (LSTM), support vector regression (SVR), extreme learning machines, and an artificial neural network with backpropagation, to define the optimal approach to predicting streamflow time series in the Carson River (California, USA) and Montmorency (Canada) catchments. The moderate resolution imaging spectroradiometer (MODIS) snow-coverage dataset was applied to improve the streamflow estimate. In addition to the DDMs, the conceptual snowmelt runoff model was applied to simulate and forecast daily streamflow. The four main predictor variables, namely snow-coverage (S-C), precipitation (P), maximum temperature (T_max), and minimum temperature (T_min), and their corresponding values for each river basin, were obtained from National Climatic Data Center and National Snow and Ice Data Center to develop the model. The most relevant predictor variable was chosen using the support vector machine-recursive feature elimination feature selection approach. The results show that incorporating the MODIS snow-coverage dataset improves the models' prediction accuracies in the snowmelt-dominated basin. SVR and LSTM exhibited the best performances (root mean square error = 8.63 and 9.80) using monthly and daily snowmelt time series, respectively. In summary, machine learning is a reliable method to forecast runoff as it can be employed in global climate forecasts that require high-volume data processing.     

An assessment of the feasibility of employing biochemical acidogenic potential tests for characterizing anaerobic biodegradability of raw and pretreated waste activated sludge

The potential to use the results of biochemical acid potential (BAP) tests to predict the ultimate digestibility of raw and pretreated waste activated sludge (WAS) was investigated. The ultimate methane production from biochemical methane potential (BMP) tests on raw and pretreated samples which spanned a range of biodegradability proved linearly related to the volatile fatty acid (VFA) and soluble chemical oxygen demand (COD) production in corresponding BAP tests. In addition, a linear relationship between NH4-N production in the BMP and BAP tests was observed. Despite the linear nature of the relationships, the ratio of the production of methane in the BMP tests to the production of VFAs in the BAP tests varied with the biodegradability of the sludge samples. Waste Activated Sludge samples with low digestibility had ultimate yields of CH4 that were greater than the VFA yields in BAP tests, whereas sludge samples with high digestibility had lower yields of CH4 than the corresponding VFA yields. This trend contrasted with the NH4 results, in which the yields in the BAP tests were consistently less than those observed in the BMP tests. It was hypothesized that the varying relationship between CH4 and VFA yields was because of the inhibition of anaerobic oxidation of long-chain fatty acids (LCFAs) in the BAP tests. Long-chain fatty acids would be converted to CH4 in BMP tests but produced as digestion intermediates in the BAP tests and were not measured as part of the VFA yield. Hydrogen and acetate were identified as the two most likely intermediates that would accumulate in the BAP tests (which would cause inhibition). A stoichiometric model to facilitate the development of an improved understanding of the biodegradation processes in the BAP and BMP tests was assembled. When the model was applied to the BAP tests the anaerobic oxidation of LCFAs and propionate and methanogenesis were excluded from the model. The model was employed to estimate the extent of degradation of lipids, carbohydrates, and proteins in the batch tests as a function of the ultimate biodegradability of the sludge samples. On the basis of model fitting, it was determined that the degradation of lipids in BMP tests decreased, whereas the degradation of carbohydrates and proteins increased as the digestibility of the sludge samples increased. The varying ratio of lipid to protein and carbohydrate degradability with increasing digestibility of the sludge samples describes the relationship between VFA production and CH4 production in the BAP, and BMP tests, respectively.