Source: PENNSYLVANIA STATE UNIVERSITY submitted to
HIGH-FREQUENCY IN-STREAM NITRATE SENSING TO ADVANCE MODELS AND INFORM WATERSHED MANAGMENT
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
EXTENDED
Funding Source
Reporting Frequency
Annual
Accession No.
1022157
Grant No.
2020-67019-31026
Project No.
PENW-2019-06490
Proposal No.
2019-06490
Multistate No.
(N/A)
Program Code
A1411
Project Start Date
Apr 15, 2020
Project End Date
Apr 14, 2025
Grant Year
2020
Project Director
Duncan, J. M.
Recipient Organization
PENNSYLVANIA STATE UNIVERSITY
408 Old Main
UNIVERSITY PARK,PA 16802-1505
Performing Department
Ecosystem Science and Manageme
Non Technical Summary
Precision conservation has the potential to dramatically reduce nonpoint source nutrient pollution. By combining high-spatial resolution geospatial data and high-frequency instream nutrient data, we will be able to develop understanding of hot spots and hot moments of nutrient pollution. Data will include high-frequency time series of nitrate concentration, discharge, dissolved oxygen, electrical conductivity, and PAR at stream reach (upstream and downstream of CSA), subcatchment, and subwatershed scales. Spatial data will include high-resolution (1m2) data of hotspots and potential BMP locations to mitigate pollution. A more systematic understanding of how multiple hydrological and biogeochemical processes control c-Q patterns across multiple spatial and temporal scales is required for effective watershed management strategies.
Animal Health Component
0%
Research Effort Categories
Basic
50%
Applied
50%
Developmental
(N/A)
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
11203992050100%
Goals / Objectives
Goal: Our project aims to develop a framework that considers the multiple processes driving temporal nitrate concentration-discharge patterns with spatial output from nutrient planning tools for more effective watershed management strategies.Objective # 1) characterize hydrological and biogeochemical drivers of concentrationdischarge variability at a range of spatial (stream reach, subcatchments, and subwatersheds) and temporal (diel, storm event, and seasonal) scales.Objective #2) determine the effects of in-stream processes on nitrate concentrationdischarge patterns over a range of flow conditions using isotope tracer and nutrient addition experiments in representative stream reaches.Objective #3) synthesize findings from objectives 1&2 in Mahantango Creek and apply foundational knowledge to guide sensor-based concentration-discharge monitoring approaches in Conewago Creek.Objective #4) leverage findings from concentration-discharge studies to better inform conservation and planning recommendations from the Agricultural Conservation Planning Framework (ACPF).
Project Methods
We will leverage high-spatial resolution land cover and topography and combine with multiple high-temporal frequency nitrate sensor data to predict nutrient contributions for different critical source areas and how they accumulate from stream reach to small watershed scales. The overall goal is to measure and deconvolve concentration-discharge (c-Q) patterns at multiple scales. The proposed project will take place at three nested scales within two HUC12 watersheds in the Susquehanna River Basin: Upper Mahantango Creek and Conewago Creek. The smallest scale is the reach scale (~100s m2) that will utilize nutrient injection experiments to calculate in-stream uptake. Reach scale measurements of dissolved oxygen, electrical conductivity, and photosynthetically active radiation will enable calculation of in-stream gross primary production, which can substantially reduce in-stream nitrate concentrations. The sub-catchment scale (~1km2) will enable quantification of concentration-discharge patterns for a small number of farms and fields. The sub-watershed scale (~10km2) will enable investigation of multiple tributaries and how proximal and distal sources of nutrients combine to comprise c-Q relationships at the outlet. Finally, the project will utilize the Agricultural Conservation Planning Framework (ACPF) GIS tool that identifies possible locations for conservation practices. The emprical data will help quantify nutrient contributions and be compared against a suite of BMPs to address nonpoint source pollution.

Progress 04/15/22 to 04/14/23

Outputs
Target Audience:The primary audience is the scientific community. We have also defined our target audience to include local stakeholders in the Conewago Creek watershed of farmers, land managers, and local government officials. Conewago Creek is a mixed land use watershed, so we needed to be deliberately broad inidentifying ourtarget audiences.We, through our partnership with Penn State's Agriculture and Environment Center,reached out to township and local government agencies as well as Susquehanna River Basin Commission to help synoptically sample water quality across the watershed seasonally. In the Mahantango Creek watershed, through our USDA-ARS colleagues, approached local land owners and farmers to gain access to farms and develop working relationships for research. This leveraged long-term presence of ARS in this watershed, which greatly helped facilitate collaborations between researchers and private land owners. Changes/Problems:Ripple effects from theCOVID pandemic continues to have impacts on this project. While offices and labs are open, there have been issues with supply chains and orders of consumables have taken longer than normal. Maintaining a continuously operating sensor network presented a substantial challenge. A notable incident involved the malfunction of the s::can sensor at FD36-3, necessitating its replacement by the manufacturer, which resulted in a five month gap in high-frequency data at this site. Additionally, a key site for our sensor network- the West Branch has added beef cattle to their field. Deployment of optical sensors in stream impacted by heavy animal agriculture is not trivial. The farmer understandably wants to keep his cattle safe and uninjured from our sensors. We have designed and tested a pump system to keep sensors and enclosures out of the stream. However, the system isn't perfect and has occasionally clogged with sediment. Therefore, large data gaps exist at multiple sites. Thus far, our sensors have survived medium sized storm events. What opportunities for training and professional development has the project provided?Training of post-doc and two undergraduate students have been facilitated through the activities of this grant. How have the results been disseminated to communities of interest? Nothing Reported What do you plan to do during the next reporting period to accomplish the goals?1) Advance data analytics including ingestion and formatting ofconcentration-dischargedata and calculation of event metrics at small (FD-36) and larger (WE-38) watershed scales. 2) Analyze data from nested high-frequency sensors in the Mahantango Creek watershed 3) Refine and employ protocols for in-stream injections to get data on in-stream nutrient uptake and spiraling. 4) Spatial locationsfrom conservation practices will be compared with c-Q patterns to determine if there is any connection.

Impacts
What was accomplished under these goals? Objective 1) Preliminary analysis of long-term (composite) concentration-discharge data was conducted at Mahantango Creek and Conewago Creek. Additionally, bi-weekly meetings were heldwith PSU and ARS collaborators to advance the analytical capabilities of high-frequency sensor data. PhD Student Melinda Marsh conducted extensive laboratory analysis of water quality samples to: a) extend and expand comparison of lab and sensor based approaches to water quality data and b) obtain water quality samples in additional locations beyond where sensors are available. . In addition to the West Branch, we continued collecting data at four additional sites within the WE-38 watershed, including FD36-3 (rock drain), FD36-1, East Branch, and the WE-38 outlet. Routine maintenance and data downloads were performed monthly for all deployed sensors, including the s::can optical sensors and the HOBO conductivity, dissolved oxygen, water level, and photosynthetically active radiation (par) sensors. We began to compile sensor data into continuous timeseries datasets for each monitoring site. Specifically, we integrated nitrate-N data from the s::can sensors with discharge measurements provided by USDA-ARS personnel to construct nitrate-N concentration-discharge timeseries. These datasets were delineated into individual storm events, and nitrate behavior was characterized using the hysteresis index and the flushing index. Each storm event was further characterized based on seasonality (Julian day, daylight hours), antecedent moisture (previous rainfall, soil moisture, water table height), and precipitation characteristics (total rainfall, duration, intensity). Objective 2) Co-PI Kennedy developed protocols for in-stream injections in Year 2 with post-doc Molly Welsh (USDA-ARS).The goal of this work is to quantify in-stream uptake of nitrate by the aquatic ecosystem including denitirification in the hyporheic zone. A multi-day field experiment in March, 2023 with multiple additions was conducted in FD-36 subwatershed. Between 4 and 5 people worked for 3 days in order to gather the necessary data and inform possible additional injections. Objective 3) The s-can sensor was maintained at the outlet of Little Conewago creek. While the site has suffered significant sediment disposition that has negatively impacted data collection, muliple storm events now have sufficient data to characterize c-Q dynamics. Objective 4) The Agricultural Conservation Planning Framework GIS tool was applied to both Conewago and Mahantango Creek watersheds. In addition to the default output, we were able to leverage another grant (USDA-CEAP program) to examine input parameter sensitivity on the number and extent of recommended conservation practices. Both watersheds also had additional field mapping and remote sensing analysis performed on conservation practice locations.

Publications

  • Type: Conference Papers and Presentations Status: Accepted Year Published: 2022 Citation: Multiscale differences of concentration-discharge relationships: implications for process inference and watershed management JM Duncan, M Marsh, AR Buda, CD Kennedy, D Millar. AGU Fall Meeting. Chicago, IL. December, 2022.
  • Type: Conference Papers and Presentations Status: Accepted Year Published: 2022 Citation: Duncan, Jonathan M; Buda, Anthony R; Kennedy, Casey D; Millar, Dave. (2022) Longer-term variations in concentration-discharge relationships. Frontiers in Hydrology Meeting, San Juan, PR
  • Type: Conference Papers and Presentations Status: Accepted Year Published: 2022 Citation: Marsh, Melinda; Duncan, Jonathan M; Buda, Anthony R; Kennedy, Casey D. (2022). Evaluating Nested Concentration-Discharge (cQ) Patterns in an Agricultural Catchment. Frontiers in Hydrology Meeting. San Juan, PR


Progress 04/15/21 to 04/14/22

Outputs
Target Audience:In Year 2 of this grant, our target audience remained roughly unchanged from Year 1. Stakeholders included farmers, land managers, and local government officials in the Conewago Creek watershed. Conewago Creek is a mixed land use watershed, so we sought to include stakeholders with interests in agricultural, forested, and urban water quality issues.Through our partnership with Penn State's Agriculture and Environment Center (led by co-PI Royer),we continued working with township and local government agencies as well as Susquehanna River Basin Commission to synoptically sample water quality across the Conewago Creek watershed on a seasonal basis. In Year 2, we expanded our outreach efforts to include new colleagues from Elizabethtown College, a nearby private university with students interested in local watershed issues. In the Mahantango Creek watershed, we continued working through our USDA-ARS partners to install sensors and monitor high-frequency nitrate-N concentrations in streams draining private farm operations in the watershed. Our intensive monitoring program would not be possible without these partnerships. Buda and Duncan presented on project plans to the USGS Chesapeake Bay Factors Group Meeting. The Factors group includes academic and government (state and federal) researchers working on water quality in the Chesapeake Bay watershed. Changes/Problems:The COVID pandemic continued to cause delays in Year 2. While some PSU and ARS policies were relaxed, there were supply chain issues that caused delays in deploying field instrumentation and analyzing water quality samples. What opportunities for training and professional development has the project provided?Research technicians at Penn State optimized the S::CAN sensor programming to maximize information content and minimize power issues. Melinda Marsh, the PhD student on the project developed protocols for laboratory analysis of water quality samples and wrote computer code in the R statistical software to analyze event-scale concentration-discharge data. How have the results been disseminated to communities of interest?The focus in Year 2 was on disseminating preliminary project findings to scientific communities. We disseminated results in two primary ways: Peer-reviewed paper in Hydrological Processes. Submitted abstracts forprofessional meeting(AGU Frontiers in Hydrology Meeting). What do you plan to do during the next reporting period to accomplish the goals?1) Deploy additional high-frequency sensors in nested fashion in the Mahantango Creek watershed. 2) Continue synoptic sampling at Conewago Creek. 3) Conduct in-stream injections in Mahantango Creek. 4) Analyze longer periods of concentration-discharge data to evaluate consistency across storm events. 5) Present preliminary results at scientific conferences to get peer feedback.

Impacts
What was accomplished under these goals? Objective 1) Preliminary analysis of long-term (composite) concentration-discharge data was conducted at Mahantango Creek and Conewago Creek. Event-scale concentration-discharge data was analyzed for multiple events at FD-36 (a subcatchment of Mahantango Creek) and WE-38 (the outlet of the long-term ARS catchment). PhD Student Melinda Marsh was recruited and began in Fall 2021. Bi-weekly meetings continued between PSU and ARS project leads. We completed the development and testing of anopen-source software package that ingests and processes large datasetsenabling thecalculation ofmetrics that describe event scale concentration-discharge patterns. This software package, which was identified in Year 1 as a high-priority need for the project, was published as a Scientific Briefing in Hydrological Processes (see Products). Objective 2) Co-PI Kennedy developed an initial protocol for in-stream injections in Year 2.This protocol will be tested and refined at USDA-ARS facilities in Massachusetts before being implemented in Pennsylvania. Objective 3) Nothing to report yet. This is scheduled to be in Year 3 of the grant Objective 4) The Agricultural Conservation Planning Framework GIS output was generated in Year 1. This output was vetted with NRCS-PA personnel in the Conewago Creek watershed.

Publications

  • Type: Journal Articles Status: Published Year Published: 2022 Citation: Millar, D.J., A.R. Buda, C.D. Kennedy, and J.M. Duncan. (2022). An open-source automated workflow to delineate storm events and evaluate concentration-discharge relationships. Hydrological Processes, 36(1), e14456.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2021 Citation: Buda, A. and Duncan, J., (August 27, 2021). High-frequency nitrate sensing to inform watershed science and management. Presented on project plans to the USGS Chesapeake Bay Factors Group Meeting. The Factors group includes academic and government researchers working on water quality in the Chesapeake Bay watershed. Virtual Meeting.


Progress 04/15/20 to 04/14/21

Outputs
Target Audience:In Year 1 of this grant, our target audience included local stakeholders in the Conewago Creek watershed of farmers, land managers, and local government officials. Conewago Creek is a mixed land use watershed, so we needed to be deliberately broad inidentifying ourtarget audiences.Through our partnership with Penn State's Agriculture and Environment Center (led by co-PI Royer), wereached out to township and local government agencies as well as the Susquehanna River Basin Commission to help synoptically sample water quality across the watershed seasonally. In the Mahantango Creek watershed, our USDA-ARS colleagues helped us approach local land owners and farmers about gaining access to farms and developing working relationships for research. This effort leveraged the long-term presence of ARS in the Mahantango Creek watershed, which greatly helped facilitate collaborations between researchers and private land owners. Changes/Problems:The COVID pandemic hadprofound impacts on this project. USDA-ARS and Penn State protocols greatly impacted our ability to make progress on this grant. Field work: travel restrictions severely limited the ability to drive in pairs to field sites. ARS was unable to travel for the first several months of the project and only for mission critical travel. Per PI Duncan's lab and department safety protocols, Penn State did not allow multiple people to travel in the same vehicle. Lab work: Water quality laboratories at Penn State and ARS were shuttered. Graduate student recruitment: We were unable to recruit a graduate student during this reporting period, In part, this was a function of the timing of funding relative to the normal academic recruiting schedule. In addition, it proved too difficult to recruit students without the ability for on-campus visits and discussions. What opportunities for training and professional development has the project provided?Research technicians at Penn State delved into the S::CAN sensor capabilities to learn how to adapt and program them for high-frequency data. How have the results been disseminated to communities of interest? Nothing Reported What do you plan to do during the next reporting period to accomplish the goals?1) Advance data analytics including ingestion and formatting ofconcentration-dischargedata and calculation of event metrics. 2) Recruit a graduate student at Penn State. 3) Deploy additional high-frequency sensors in nested fashion in the Mahantango Creek watershed. 4) Design protocols for in-stream injections.

Impacts
What was accomplished under these goals? Objective 1) Preliminary analysis of long-term (composite) concentration-discharge data was conducted at Mahantango Creek and Conewago Creek. Additionally, bi-weekly meetings were heldwith PSU and ARS collaborators to advance the analytical capabilities of high-frequency sensor data. Through these meetings, we identified theneed for an open-source software package to ingest and process large datasetsenabling thecalculation ofmetrics that describe event-scale concentration-discharge patterns. Objective 2) Nothing to report yet. In collaboration with Co-PI Kennedy, we will develop protocols for in-stream injections in Year 2. Objective 3) Nothing to report yet. This is scheduled to be in Year 3 of the grant Objective 4) The Agricultural Conservation Planning Framework (ACPF) GIS tool was applied to both Conewago and Mahantango Creek watersheds. In addition to the default ACPF output, we were able to leverage another grant (funded by the USDA-CEAP program) to examine input parameter sensitivity on the number and extent of recommended conservation practices.

Publications