14-1 Stream Widening and Evolving Flood Risk Driven by Land-Use and Climate Changes in Eastern Pennsylvania Using Field, Remote Sensing, and Modeling Approaches

Session: Lake Sedimentary Records of Past Climate and Environment

Presenting Author:

Authors:

Abstract:

This study seeks to apportion land use and precipitation changes in driving the apparent increase in the number and scale of damaging floods in the mid-Atlantic region. Sedimentologic, channel widening, and discharge time series are reconstructed with field, remote sensed, and numerical modeling approaches using the Saucon watershed in eastern Pennsylvania as a natural laboratory. The Saucon watershed offers a decadal-scale overlapping record of channel widening in air photos and annual resolution of sediment transport and deposition preserved in mill pond deposits that can be tested against a Landlab-based numeric model that varies hydrologic parameters, proxies for changes in infiltration and precipitation. The channel widening record, using a collection of 90-year air photos, indicates channel narrowing over time. Following a base-level fall, the channel experienced narrowing into the 1970s. From 1980 to the present, the channel exhibits alternating phases of narrowing and widening. Using supervised classifications in ArcGIS, land use is shown to have transitioned from predominantly agricultural cover toward reduced farmland and increased forested areas. The percentage of land use collected from the classifications will determine model parameters. The mill pond record is 4 m thick and composed of 40 cm of Holocene colluvium overlain by 50 cm of fluvial sandy gravel, 30 cm of organic-rich sandy overbank/marsh, and 280 cm of deltaic and lacustrine pebbly sand and mud. A basal maximum calibrated ¹⁴C age of 1602 calendar years indicates 0.93 cm/yr. A magnetic susceptibility time series has a concentration of power at 8.5 cm, mimicking visible fining-up cycles in the lacustrine facies that we interpret as a decadal-scale variation in floods inundating the pond. These results are compared to the predictions of a 5-component Landlab model that varies infiltration rates, precipitation duration, and intensity. The purpose of the model is to evaluate watershed response to varied precipitation and infiltration rates and to determine whether these forcings reproduce decadal-scale flood signs comparable to those observed in the mill pond. This research furthers understanding of how watersheds respond to changing land use and precipitation and informs sustainable land and water management decisions targeting vulnerable communities' adaptation to future flood risks.

© Copyright 2025 The Geological Society of America (GSA), all rights reserved.