Radiocarbon data show that a fraction of riverine organic carbon (OC) has been pre-aged in the terrestrial environment. Relatively little is known, however, about the regional climatic, anthropogenic, and landscape factors that promote the mobilization of aged OC to rivers. This study examines associations between riverine OC and river basin characteristics. It leverages data from three sources: 1) a spatially extensive compliation of literature-reported radiocarbon measurements from the conterminous United States, 2) the U.S. Environmental Protection Agency’s Stream-Catchment (StreamCat) database and 3) fire regime data from the U.S. Department of Agriculture Forest Service Landscape Fire and Resource Management Planning Tools (LANDFIRE, 2021).  The compiled radiocarbon data include 99 dissolved organic carbon (DOC) and 57 particulate organic carbon (POC) ages, respectively after averaging by location. We used the random forest (RF) machine learning algorithm to build independent models with StreamCat data as predictor variables.  Each model used 3,000 trees and a random 75:25 data split for model validation, and results were averaged across 1,000 iterations (mean 14C-DOC-MSE = 5,186.1, r = 0.54; mean 14C-POC-MSE = 12,299.8, r = 0.68). Our results suggest that the controls on radiocarbon age vary between the OC pools. Anthropogenic drivers were associated with the presence of old POC while land cover and related processes were the primary variables affecting DOC age.

Primary Presenter: S. Leigh McCallister, Virginia Commonwealth University (slmccalliste@vcu.edu)

Authors:

Kaycee Faunce, USGS  (kfaunce@usgs.gov)

Daniel McGarvey, Virginia Commonwealth University  (djmcgarvey@vcu.edu)

S. Leigh McCallister, Virginia Commonwealth University  (slmccalliste@vcu.edu)