Ungaged inflow and loss patterns in urban and agricultural sub‐reaches of the Logan River Observatory

Abstract

Streams in semi-arid urban and agricultural environments are often heavily diverted for anthropogenic purposes. However, they simultaneously receive substantial inflows from a variety of ungaged sources including stormwater returns, tile drainage, and irrigation runoff that help sustain flow during dry periods. Due to the inability to identify sources or directly gage many of these inflows, there is a clear need for methods to understand source origination while quantifying potential gains and losses over highly impacted reaches. In the context of the Logan River Observatory, historical gage data illustrate the importance of ungaged and unidentified inflows on maintaining or enhancing flows in both urban and agricultural reaches containing large diversions. To understand the inflows in this portion of the Logan River, we first analysed water samples for ions collected from a subset of representative inflow sources and applied clustering analyses to establish inflow source classifications and associated ion concentration ranges. These representative concentration ranges, combined with mainstem flow and river ion samples taken at sub-reach scales, allow for the application of flow and mass balances to quantify inflow rates from different sources as well as any losses. These calculations demonstrate significant gains and losses occurring in many sub-reaches during three sampling events. The dominant land use (urban or agriculture) and flow regime at the time of sampling were the primary drivers of gains and losses. These exchanges were found to be most important below large diversions during low flow conditions. This highlights the need to classify inflow sources (urban or agriculture, surface or groundwater) and estimate their contributions to anticipate instream consequences of land use and water management decisions. As irrigation and water conveyance practices become more efficient, a portion of these ungaged inflows could be diminished or eliminated, thus further depleting streamflow during dry periods.

The Role of Inflow and Infiltration (I/I) in Urban Water Balances and Streamflow Regimes: A Hydrograph Analysis Along the Sewershed‐Watershed Continuum

“…Such unaccounted for fluxes are common in urban areas throughout the United States (Li et al., 2020). Additional sources of I/I may include net decline of stored groundwater or unaccounted water sources within the catchment such as leakage from water distribution systems and urban irrigation (Bhaskar & Welty, 2012; Bhaskar et al., 2016; Fillo et al., 2021; Lancia et al., 2020; Lerner, 2002; Sanzana et al., 2019; Tennant et al., 2021). There is evidence showing that the groundwater table in urban Milwaukee is dropping over time in recent years (Choi et al., 2012).…”

Machine learning for hydrologic sciences: An introductory overview

“…Over the past few decades, variants of K-means and other algorithms such as agglomerative hierarchical clustering and fuzzy clustering have been proposed and used in various applications (de Oliveira & Pedrycz, 2007;Jain, 2010;Murtagh & Legendre, 2014;H. Tennant et al, 2021).…”

“…Over the past few decades, variants of K‐means and other algorithms such as agglomerative hierarchical clustering and fuzzy clustering have been proposed and used in various applications (de Oliveira & Pedrycz, 2007; Jain, 2010; Murtagh & Legendre, 2014; H. Tennant et al, 2021). Although clustering is an unsupervised learning technique, it is sometimes used to learn data representation in the pre‐processing step for a supervised learning task.…”

Detailed streamflow data for understanding hydrologic responses in the Logan River Observatory

“…However, the 2017 flow balance calculations for this reach show a large gain in water (positive %ΔQ) over this section (Figure 2d). More detailed data collection in this reach have shown that these gains can be attributed to multiple, ungaged, surface and subsurface inflows from urban irrigation returns, irrigation distribution systems, and groundwater drains (Buahin et al, 2019;Tennant et al, 2021) that vary in magnitude over time due to changing anthropogenic impacts and hydrologic conditions. These simple net flow balance calculations also provide a foundation to interpret water quality data.…”

“…However, within the lower portion of the watershed, there are also many unknown and ungaged inflows that help maintain some amount of flow in the Logan River during these periods. Tennant et al (2021) provide details regarding these influences.…”