Flow regimes are critical for determining physical and biological processes in rivers, and their classification and regionalization traditionally seeks to link patterns of flow to physiographic, climate and other information. There are many approaches to, and rationales for, catchment classification, with those focused on streamflow often seeking to relate a particular response characteristic to a physical property or climatic driver. Rationales include such topics as Prediction in Ungauged Basins (PUB), and providing guidance for model selection in poorly understood hydrological systems. The Annual Daily Hydrograph (ADH) is a first-order easily visualized integrated expression of catchment function, and over many years the average ADH is a distinct hydrological signature that differentiate catchments from each other. In this study, we use t-SNE, a state-of-the-art technique of dimensionality reduction, to classify 17110 ADHs for 304 reference catchments in mountainous Western North America. t-SNE is chosen over other conventional methods of dimensionality reduction (e.g. PCA) as it presents greater separability of ADHs, which are projected on a 2D map where the similarities are evaluated according to their map distance. We then utilize a Deep Learning encoder to upgrade the non-parametric t-SNE to a parametric approach, enhancing its capability to address ’unseen’ samples. Results showed that t-SNE successfully clustered ADHs of similar flow regimes on the 2D map and allowed more accurate classification with KNN. In addition, many compact clusters on the 2D map in the coastal Pacific Northwest suggest information redundancy in the local streamflow network. The t-SNE map provides an intuitive way to visualize the similarity of high-dimensional data of ADHs, groups catchments with like characteristics, and avoids the reliance on subjective hydrometric indicators. This article is protected by copyright. All rights reserved.

Phosphorus (P) loss in agricultural discharge has typically been associated with surface runoff; however, tile drains have been identified as a key P pathway due to preferential transport. Identifying when and where these pathways are active may establish high-risk periods and regions that are vulnerable for P loss. A synthesis of high-frequency, runoff data from eight cropped fields across the Great Lakes region of North America over a 3-yr period showed that both surface and tile flow occurred year-round, although tile flow occurred more frequently. The relative timing of surface and tile flow activation was classified into four response types to infer runoff-generation processes. Response types were found to vary with season and soil texture. In most events across all sites, tile responses preceded surface flow, whereas the occurrence of surface flow prior to tile flow was uncommon. The simultaneous activation of pathways, indicating rapid connectivity through the vadose zone, was seldom observed at the loam sites but occurred at clay sites during spring and summer. Surface flow at the loam sites was often generated as saturation-excess, a phenomenon rarely observed on the clay sites. Contrary to expectations, significant differences in P loads in tiles were not apparent under the different response types. This may be due to the frequency of the water quality sampling or may indicate that factors other than surface-tile hydrologic connectivity drive tile P concentrations. This work provides new insight into spatial and temporal differences in runoff mechanisms in tile-drained landscapes.