Location: Hydrology and Remote Sensing Laboratory
Project Number: 8042-13610-030-052-I
Project Type: Interagency Reimbursable Agreement
Start Date: Oct 1, 2021
End Date: Sep 30, 2024
Objective:
Wetlands in agricultural landscapes provide vital ecosystem services but quantification of ecosystem service provision is often lacking. This project will quantify the importance of the ecosystem services associated with natural and restored wetlands in the Mid Atlantic Region.
[1] Characterize ecosystem service provisioning for soil organic carbon in protected and restored wetlands and riparian ecosystems.
[2] Assessment of pollination ecosystem service provisioning by wetland conservation practice easements.
[3] Assess agricultural nitrate mitigation potential using MESA as an indicator of cropland influenced water.
[4] Quantifying hydrologic change with wetland restoration using time-series Radarsat-2 SAR data.
[5] Wetland hydroperiods in Delmarva using time-series SAR data.
Approach:
Obj. 1: ARS will assess the variability and capacity of wetland carbon storage through land use change by characterizing the ecosystem service provisioning by protected natural, restored, and drained wetlands and with a longer-term goal of evaluating the validity of the SWAT-C model. ARS will use empirical field data to calculate and aggregate soil organic carbon stocks for natural, restored, and drained wetlands and extrapolated to riparian areas located in USDA program easements in the Chesapeake Watershed.
Obj. 2: ARS will use the National Capital Project’s InVEST (Integrated Valuation of Ecosystem Services and Tradeoffs) model for crop pollination to determine the effectiveness of non-traditional aquatic protected lands to provide habitat for pollinator populations and services within our developed database of protected conservation lands in the Chesapeake Watershed.
Obj. 3 Using MESA as a tracer, this study will assess the extent to which different wetlands can intercept groundwater influenced by corn/soy cropland and assess likelihood that wetlands are processing agricultural nitrate based on ecosystem type and location. This information can then inform ecosystem models estimating nitrate processing by wetlands on the Delmarva Peninsula.
Obj. 4: With the combination of Radarsat-2 SAR data and in-situ data, spatial maps of inundation and soil moisture will be generated to assess effects of a wetland restoration project on hydrologic change at ~2,000-acre forested Wetland Reserve Program site in Somerset County, Maryland. Deep Neural Network (DNN) methods will be used to establish the non-linear relationship between SAR backscatter coefficients and inundation status and soil moisture.
Obj. 5: The aim of this study is to evaluate an innovative Convolutional Neural Network (CNN) in combination with artificial recurrent neural network (RNN) for wetland hydroperiods mapping and analysis based on time series Sentinel-1 C band SAR in Delmarva in most recent years.
