Files
Abstract
When modeling wet weather flow in sanitary sewers the RTK method is often used. The RTK
method converts a specified amount of rainfall into sewer flow, regardless of changes in the
soil’s moisture content and saturation. Robert Czachorski and Tobin VanPelt [2001] sought to
remedy this issue when they developed the antecedent moisture modeling [AMM] procedure,
which explicitly accounts for antecedent moisture conditions. AMM tracks antecedent moisture
using rain and temperature signals. The rain signal adds moisture to the soil according to the
magnitude of the temperature signal. High temperatures reduce added soil moisture because
evaporation and transpiration are higher, resulting in drier antecedent moisture conditions, while
low temperatures do the reverse. Since 2021, AMM has been applied with great success on
hundreds of projects [Edgren et al., 2023]. AMM, in its current form as documented by
Czachorski, Edgren and Gonwa [2023], has the potential to revolutionize sanitary sewer system
hydrologic modeling. Nearby streams experience roughly the same antecedent moisture
conditions as a sanitary sewer system. Is it possible that streamflow provides a better indicator
signal of antecedent moisture conditions than temperature? The purpose of this paper is to
explain an investigation with a focus on determining if streamflow provides a better indicator of
antecedent moisture conditions than temperature. AMM using temperature was compared against
AMM using streamflow, and the models were calibrated using a Bayesian Optimization
algorithm. The results indicated that streamflow produces a comparable or better representation
of antecedent moisture conditions. Out of the various streamflow methods, two-level models
utilizing a large regional stream produced the lowest error. Future studies investigating
streamflow’s ability to represent antecedent moisture should include a larger sample size of
sewer systems in order to validate these results.