MODELING PHOSPHORUS LOADING TO THE CANNONSVILLE RESERVOIR USING SWAT

Similar documents
Nitrate Load Reduction Strategies for the Raccoon and Des Moines Rivers. Keith Schilling, Calvin Wolter Iowa DNR Geological and Water Survey

Overview of the Surface Hydrology of Hawai i Watersheds. Ali Fares Associate Professor of Hydrology NREM-CTAHR

Dairy Outlook. January By Jim Dunn Professor of Agricultural Economics, Penn State University. Market Psychology

An Environmental Accounting System to Track Nonpoint Source Phosphorus Pollution in the Lake Champlain Basin. Year 2 Project Work Plan

Utilization of the SWAT Model and Remote Sensing to Demonstrate the Effects of Shrub Encroachment on a Small Watershed

Cattle Outlook. January, 2018

Joint Research Centre (JRC)

The Fourth Assessment of the Intergovernmental

Economic and Phosphorus-Related Effects of Precision Feeding and Forage Management at a Farm Scale

Dairy Outlook. April By Jim Dunn Professor of Agricultural Economics, Penn State University. Market Psychology

Lecture 9A: Drainage Basins

The Impacts of Climate Change on Portland s Water Supply

CALS Integrated Nutrient Management for Dairy and Livestock Farms PWT

Dairy Outlook. June By Jim Dunn Professor of Agricultural Economics, Penn State University. Market Psychology

Grazing Management Different Strategies. Dr Jim Russell and Joe Sellers Iowa State University

Application of a Basin Scale Hydrological Model for Characterizing flow and Drought Trend

M.L. Kavvas, Z. Q. Chen, M. Anderson, L. Liang, N. Ohara Hydrologic Research Laboratory, Civil and Environmental Engineering, UC Davis

Modeling Nutrient and Sediment Losses from Cropland D. J. Mulla Dept. Soil, Water, & Climate University of Minnesota

San Antonio Water System Mitchell Lake Constructed Wetlands Below the Dam Preliminary Hydrologic Analysis

Hood River Water Conservation Strategy: achieving long-term water resource reliability for agriculture & local fish populations

Development of a tool to estimate Best Management Practices (BMP) efficiency using SWAT

History of Model Development at Temple, Texas. J. R. Williams and J. G. Arnold

Surface Soil Moisture Assimilation with SWAT

Milk Production. January Milk Production up 2.7 Percent

Overview of Models for Estimating Pollutant Loads & Reductions

Fertility management in organic strawberries

The Florida Ranchlands Environmental Services Project (FRESP)

Rainwater Management. Dr. Iftikhar Ahmad. College of Earth and. University of The Punjab Lahore

2017 Tennessee Agricultural Outlook. Aaron Smith Crop Economist University of Tennessee Extension

Cattle Market Situation and Outlook

Climate Change & Urbanization Have Changed River Flows in Ontario

FORAGE SYSTEMS TO REDUCE THE WINTER FEEDING PERIOD. Gerald W. Evers

Nutrient Budgeting. An Overview of What, How and Why. June 2014

Rock Creek Floodplain Analysis

Coupling SWAT with land cover and hydropower models for sustainable development in the Mekong Basin

USING ARCSWAT TO EVALUATE EFFECTS OF LAND USE CHANGE ON WATER QUALITY. Adam Gold Geog 591

Comparison of Statistical and Dynamic Downscaling methods for Hydrologic Applications in West Central Florida

Administration Division Public Works Department Anchorage: Performance. Value. Results.

Information Request 11

Implementation timeline for the Water Framework Directive

SOIL P-INDEXES: MINIMIZING PHOSPHORUS LOSS. D. Beegle, J. Weld, P. Kleinman, A. Collick, T. Veith, Penn State & USDA-ARS

NBI strategic water resources analysis Phase I findings

Arkansas Water Resources Center

Uncertainty in hydrologic impacts of climate change: A California case study

Carbon Sequestration in California s Rangeland Soils

Water Quality Study In the Streams of Flint Creek and Flint River Watersheds For TMDL Development

Cattle & Beef Outlook

CHAPTER FIVE Runoff. Engineering Hydrology (ECIV 4323) Instructors: Dr. Yunes Mogheir Dr. Ramadan Al Khatib. Overland flow interflow

Feed Grain Outlook June 2, 2014 Volume 23, Number 33

Nutrient BMP Evaluation Project: Preliminary Conclusions

Nutrients and Ecosystems

Impact of Point Rainfall Data Uncertainties on SWAT Simulations

Hydrologic Modeling with the Distributed-Hydrology- Soils- Vegetation Model (DHSVM)

SAN BERNARD RIVER WATER QUALITY MODEL UPDATE. August 18, 2011

EVALUATION OF HYDROLOGIC AND WATER RESOURCES RESPONSE TO METEOROLOGICAL DROUGHT IN THESSALY, GREECE

Cattle Situation and Outlook

Central Texas Cow/Calf Clinic

SURFACE WATER WITHDRAWALS & LOW FLOW PROTECTION POLICY MICHAEL COLLEGE, P.E. SUSQUEHANNA RIVER BASIN COMMISSION

WATER RESOURCES MANAGEMENT Vol. II - Watershed Modeling For Water Resource Management - D. K. Borah WATERSHED MODELING FOR WATER RESOURCE MANAGEMENT

MONITORING HEIFER PROGRAMS

Veal Price Forecast. October 2015

Use of SWAT for Urban Water Management Projects in Texas

PENNSYLVANIA PHOSPHORUS INDEX UPDATE

Regionalization of SWAT Model Parameters for Use in Ungauged Watersheds

I/I Analysis & Water Balance Modelling. Presented by Paul Edwards

Urbanization effects on the hydrology of the Atlanta area, Georgia (USA)

Historical Prices, Trends, Seasonal Indexes, and Future Basis of Cattle and Calves at Sioux Falls, South Dakota

Pete Fandel Illinois Central College llinois Council on Best Management Practices

Situation and Outlook of the Canadian Livestock Industry

Does Water Resources Management in the Snake River Basin Matter for the Lower Columbia River? Or Is the Snake River Part of Another Watershed?

EMERGING ISSUES WITH ALFALFA AND FORAGES IN IDAHO

Using Paired Edge of Field Data to Assess Impacts of Management on Surface and Subsurface P Loss

global science solutions

Estimation of Actual Evapotranspiration at Regional Annual scale using SWAT

Appendix 12. Pollutant Load Estimates and Reductions

Pennsylvania s Phase III Watershed Implementation Plan

Livestock and Feedgrain Outlook

TEXAS A8cM UNIVERSITY TEXAS AGRICULTURAL EXTENSION SERVICE

Application of SWAT Model in land-use. change in the Nile River Basin: A Review

IMPACT OF CLIMATE CHANGE ON WATER AVAILABILITY AND EXTREME FLOWS IN ADDIS ABABA

DRAFT Great Bay Nitrogen Non-Point Source Study

Eco-hydrological assessment of German lowland catchments experiences and challenges- N. Fohrer, B. Schmalz, G. HörmannH

Electric Forward Market Report

Profitability of Tasmanian beef enterprises:

Assessing Climate Change Impacts on Water Resources in the Beas Basin &

MODELING LONG-TERM WATER QUALITY IMPACT OF STRUCTURAL BMPS

Soil Temperature Damping Depth in Boreal Plain Forest Stands and Clear Cuts: Comparison of Measured Depths versus Predicted based upon SWAT Algorithms

Beef Cattle Market Outlook

U.S. Packing Capacity Sufficient for Expanding Cattle Herd

The Confluence Model. Presentation to Modeling and Forecasting Working Group January 21, 2015

MODELING SEDIMENT AND PHOSPHORUS YIELDS USING THE HSPF MODEL IN THE DEEP HOLLOW WATERSHED, MISSISSIPPI

MODELING A SMALL, NORTHEASTERN WATERSHED

July, International SWAT Conference & Workshops

What tools can we use to help us decide when to enter and when to exit a hedge? (Or, or for that matter, when to enter and exit any trade.

An investment model for large-scale green

The Vermont Dairy Farm Sustainability Project, Inc.

12/28/2016. Air. Surface Water. Ground Water. Soil. 1. Calculate agronomic rate. 2. Identify optimal fields. 3. Determine when to apply

Leila Talebi and Robert Pitt. Department of Civil, Construction, and Environmental Engineering, The University of Alabama, P.O. Box , Tuscaloosa

Longitudinal studies refer to those that gather information from the same set of respondents through repeated visits over a defined period of time.

Livestock and Poultry Environmental Learning Center Webcast Series June 20, From: G. Albrecht P. Ristow

Transcription:

MODELING PHOSPHORUS LOADING TO THE CANNONSVILLE RESERVOIR USING SWAT Bryan Tolson 1 & Christine Shoemaker 2 1. PhD Student, 2. Professor School of Civil & Environmental Engineering Cornell University PWT Meeting

Why are We Here? To get feedback on our results for future projections, including suggestions for improvement. To discuss our current modeling results

Study Area Cannonsville Reservoir Basin agricultural basin New York City water supply P restriction impedes economic growth of county County wishes to investigate the impact of different management practices on P loading to reservoir Develop Watershed Model

Highly Significant Results Our model results show that phosphorous loading to the reservoir will INCREASE over the years, even if the cow population stays at its currently low (1997) level. This is a politically sensitive result so we want to be sure that this trend represents a reasonable prediction before we discuss it publicly. We would appreciate your not discussing these results widely at this time until we have had time to incorporate your comments.

Informal Talk with Preliminary Results Because we wanted your feedback, we are asking your opinion early in the process of evaluating future scenarios. As a result, many of the results you will see are preliminary.

Status of Modeling Work Refining model calibration with historical data based on peer review of modeling report. (This work is quite polished.) Initial simulations of future P loading to the reservoir. (This work is preliminary.)

Predictions of Future P Loading What do we think will happen to total P loading to the reservoir? Initially, we simulated the model with constant present day (1997) P inputs for 40 years for two cases: With previous 40 years of historical climate data With steady-state climate inputs (repeat climate)

Long-Term Simulation Results Simulation using previous 40 years of climate 40000.0 35000.0 30000.0 25000.0 20000.0 15000.0 10000.0 Total P to Reservoir 2007 2011 2015 2019 2023 Annual Total P load 2027 2031 2035 2039 2043 Year Over 37 yrs, Total P load increases by about 40% why? Climate or P inputs to basin? apparently climate as flows and sediment load increase similarly

Long-Term Simulation Results Simulation results under steady-state climate inputs (1994-95 climate is repeated) Annual Total P load in kg 55000 50000 45000 40000 35000 30000 Total P to Reservoir Over 37 yrs, Total P load increases by about 7% flow and sediment increase by less than 1% in this period 2004 2008 2012 2016 2020 Year 2024 2028 2032 2036 2040

Long-Term Simulation Results Two immediate questions: 1. Why are these results significant? 2. Are these simulation results reasonable?

Long-Term Simulation Results - Significance If conditions do not worsen in the basin, loading could approach TMDL over time The latest monitoring data show a decrease in total P loading over the past two years such that P restriction is not in effect NYC DEP modeling work with GWLF concluded that they do not expect to see a P loading increase in the foreseeable future

Alternate Approach to SWAT Model: Basin-Wide Mass Balance The SWAT model makes many assumptions and gives results that are spatially and temporally distributed. The mass balance looks only at annual inputs and outputs of phosphorous to the watershed. The goal is to see whether the mass balance would support the SWAT results that there is a build up of phosphorous in output over time.

Current Basin-Wide P Mass Balance P INPUTS in kg/yr 20 19 18 11 12 13 Net dairy farm P= +142,500 to +307,500 Atmospheric P deposition = 5,000 Septic System P = 6,300 WWTP P = 4,000 TOTAL above = 158,000 to 323,000 Other inputs not in above? E.g. beef farms, other agricultural 43 38 10 9 37 8 7 42 36 35 26 6 34 5 4 33 27 3 32 31 2 30 28 1 41 NET P ACCUMULATION is 107,000 to 272,000 kg/yr in basin Inaccuracies seem to be negligible compared to magnitude of inputs. 21 22 17 29 40 39 16 24 14 25 15 23 P loaded to reservoir: 51,400 kg/yr Cash crop harvest:? OUTPUTS Lumber P harvest:?

Long-Term Simulation Results Realistic? Purpose of our presentation is to gain your feedback to determine if results are realistic: 1. Do the available P mass balance data support this finding? 2. Are we modeling the P inputs, outputs and cycling in the system accurately enough to be relatively confident in such a finding?

Modeling of Cannonsville Reservoir Basin Soil and Water Assessment Tool version 2000 (SWAT2000) Continuous-time, spatially distributed model Not designed for detailed, single-event flood routing SWAT2000 has an ArcView 3.2 Interface (AVSWAT) to process input data and create model inputs Developed by USDA and Texas A&M Primarily agricultural model Applied throughout the US, including Texas, Illinois, and Pennsylvania

Building Block of SWAT: The Hydrologic Response Unit (HRU) Land Use + Soils Hydrologic Response Unit Model subdivides basin into subbasins Land use and soils are intersected within each subbasin These intersections are modeled independently within each subwatershed Average HRU size in this application is about 2 km 2

AVSWAT Processing Results HRUs & Subbasins in this application 3 2 1 20 21 22 19 17 29 18 16 15 11 12 13 40 14 25 23 39 24 43 38 10 9 37 6 7 8 34 42 36 35 26 5 4 33 27 32 31 30 28 41 Note: HRUs are not spatially continuous, thus constituents are only routed at the subbasin scale

The Concept Behind SWAT Time/location Land Use + Soils Hydrologic Response Unit (HRU) Time routing Land Management Practices Streamflow Groundwater Surface Water Evapotranspiration Suspended Sediment (Soil) Phosphorus Concentrations

Soil Phosphorus Cycle in SWAT P forms tracked in SWAT

P-Source Representation in the SWAT Basin Model Manure and fertilizer P Forest P Groundwater and Septic System P Urban stormwater P Point source P (WWTP s)

Manure P Inputs to SWAT Manure mass, P content and timing of application Manure representation is approached from a mass balance basis at the subbasin level Subbasin specific cattle population data are used to estimate manure production rates Subbasin manure loads then allocated to the available agricultural land covers Dairy 90% of cattle manure P Beef 10% of cattle manure P Daily spreading is assumed

Spatial Manure Production and Application 1997 estimates of the cattle population Manure application timing and masses vary by land cover in each subbasin 3 2 1 20 21 22 19 17 29 11 18 12 13 39 40 16 24 14 25 15 23 43 10 38 6 7 9 8 42 34 35 36 37 26 5 4 32 33 27 31 30 41 28 Rivers Subbasin Manure P in kg/ha 0-0.1 0.1-0.4 0.4-1.3 1.3-2.5 2.5-3.7 3.7-6.1 Sources: P. Cerosaletti, D. Dewing, K. Czymmek, 1992 NRCS Ag. Survey in Cannonsville basin, US Ag. Survey

Model Development: Calibration and Validation Model calibrated from Jan. 1994 to Sept. 2000 for flow, sediment and phosphorus Model validated from Jan. 1990 to Dec. 1993 for hydrology and from Oct. 1991 to Dec 1993 for sediment and phosphorus Main calibration and validation water quality station drains about 80% of the basin

Model Results Hydrology 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Flow @ Walton Measured Predicted CALIBRATION Statistics: R 2 = 0.73 % Difference: 4.1 Jan-94 Apr-94 Jul-94 Oct-94 Jan-95 Apr-95 Jul-95 Oct-95 Jan-96 Apr-96 Jul-96 Oct-96 Jan-97 Apr-97 Jul-97 Oct-97 Jan-98 Apr-98 Jul-98 Oct-98 Jan-99 Apr-99 Jul-99 Oct-99 Jan-00 Apr-00 Jul-00 Avg monthly flow in CMS Date 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Flow @ Walton Jan-90 Mar-90 May-90 Jul-90 Sep-90 Nov-90 Jan-91 Mar-91 May-91 Jul-91 Sep-91 Nov-91 Jan-92 Mar-92 May-92 Jul-92 Sep-92 Nov-92 Jan-93 Mar-93 May-93 Jul-93 Sep-93 Nov-93 Avg monthly flow in CMS VALIDATION Statistics: R 2 = 0.83 % Difference: 10.0 Measured Predicted Date

Model Results Total Phosphorus 30000.0 25000.0 20000.0 15000.0 10000.0 5000.0 TOTAL P (TP) Loading @ Beerston *98,000 Measured Predicted CALIBRATION Excluding Jan 96 R 2 : 0.47 % Difference: 4.3 0.0 Jan-94 Apr-94 Jul-94 Oct-94 Jan-95 Apr-95 Jul-95 Oct-95 Jan-96 Apr-96 Jul-96 Oct-96 Jan-97 Apr-97 Jul-97 Oct-97 Jan-98 Apr-98 Jul-98 Oct-98 Jan-99 Apr-99 Jul-99 Oct-99 Jan-00 Apr-00 Jul-00 Monthly TP Load (kg) Date VALIDATION R 2 = 0.59 % Difference: 1.0 30000 25000 20000 15000 10000 5000 0 Total P (TP) Loading @ Beerston Measured Predicted Oct-91 Nov-91 Dec-91 Jan-92 Feb-92 Mar-92 Apr-92 May-92 Jun-92 Jul-92 Aug-92 Sep-92 Oct-92 Nov-92 Dec-92 Jan-93 Feb-93 Mar-93 Apr-93 May-93 Jun-93 Jul-93 Aug-93 Sep-93 Oct-93 Nov-93 Dec-93 Monthly TP Load (kg) Date

Long-Term Simulation Results Realistic? Annual Total P load in kg 55000 50000 45000 40000 35000 30000 Total P to Reservoir 2004 2008 2012 2016 2020 Year 1. Do the available P mass balance data support this finding? 2. Are we modeling the P inputs, outputs and cycling in the system accurately enough to be relatively confident in such a finding? 2024 2028 2032 2036 2040

Model Check kg of P to reservoir 50000 45000 40000 35000 30000 25000 20000 15000 10000 5000 0 21 % reduction in 2004 Total P to Reservoir No Ag P Added to Soil Ag P Added to Soil 45 % reduction in 2077 2004 2008 2012 2016 2020 2024 2028 2032 2036 2040 2044 2048 2052 2056 2060 2064 2068 2072 2076 Year - two-year repeating climate, steady-state inputs

1. Current Basin-Wide P Mass Balance P INPUTS in kg/yr 20 19 18 11 12 13 Net dairy farm P= +142,500 to +307,500 Atmospheric P deposition = 5,000 Septic System P = 6,300 WWTP P = 4,000 TOTAL above = 158,000 to 323,000 Other inputs not in above? E.g. beef farms, other agricultural 43 38 10 9 37 8 7 42 36 35 26 6 34 5 4 33 27 3 32 31 2 30 28 1 41 NET P ACCUMULATION is 107,000 to 272,000 kg/yr in basin Inaccuracies seem to be negligible compared to magnitude of inputs. 21 22 17 29 40 39 16 24 14 25 15 23 P loaded to reservoir: 51,400 kg/yr Cash crop harvest:? OUTPUTS Lumber P harvest:?

1. Details of Basin P Mass Balance P INPUTS TO WATERSHED: Septic Systems (Day, 2001) WWTP s monitoring data Atmospheric deposition P budget study of Hubbard Brook Experimental Forest (Yanai, 1992). Estimated atmospheric P inputs in rainfall at 0.04 kg P/ha/yr Other studies suggest atmospheric inputs > 0.04 kg P/ha/yr NET basin-wide dairy farm P mass balance extrapolated from dairy cattle population estimates and dairy P mass balance studies [Klausner (1993) & Tylutki & Fox (1997)] as reported in Cerosaletti (2002)

1. Details of Basin P Mass Balance NET DAIRY P INPUTS TO WATERSHED: data from previous sources suggest P accumulation rates at the farm scale range from 19 to 41 kg P/per cow/yr our current estimates put the basin-wide dairy cow population at 7500 cows (not counting calves, heifers) thus, extrapolating to basin-scale we estimate: 19*7500 to 41*7500 OR 142,500 to 307,500 kg P/yr remaining on dairy farms in the basin. even if the number of cows was 5000, this interval is still 95,000 to 205,000 kg P/yr

1. Question Assuming that net P inputs to the basin are greater than net P outputs, then: Where is the net increase in P on dairy farms going? Model represents the net increase by either increasing plant uptake or continually adding P to second soil layer.

1. P TRANSPORT WITH RESPECT TO SOIL LAYER 2 AS MODELED IN SWAT Rainwater Soluble Reactive P concentration Fertilizer = 0 P Top 10 mm (1) Layer Adjacent to Top 10 mm '(2) Surface Runoff interacting with layer 1 to transport all forms of P Leaching of soluble P to the second soil layer Remaining Layers (3) No transport of P modeled in layer 2 except plant uptake. no P flux across thick black boundary

1. Soil Profile P Long-term, does this continual build-up of soil P in soil layer 2 represent the realistic accumulation of soil P? In reality is there significant losses of soil P over the long-term from the second soil layer thru leaching or lateral transport? OR, does the P accumulate in such a way that it is totally unavailable for transport to streams?

2. Simplified P Cycle on Dairy Farm Based on Cerosaletti (2002) Purchased Feed COW Milk currently, we do not explicitly model these processes. CROPS SOILS MANURE We do simulate the harvest of crops and the generation of manure from a given cattle population. Exit soil profile Fertilizer Farm Boundary

2. Model Representation of P Cycling Net dairy farm P Net P = P in manure + P in Inorganic Fert. - P in removed crops 214,000 kg P in dairy manure + 25,800 kg inorganic P fertilizer spread on fields How much of this manure is part of the dairy farm P cycle? Corn Silage, Hay and pasture currently removed from HRUs by harvest or grazing (assumed to become feed in basin) Biomass grazed in pasture = approx. estimate of dry biomass consumed during grazing by current cattle population. Checking that total dry biomass harvested or grazed is less than dry biomass intake of current cattle population.

2. Model Representation of P Cycling If harvest/graze of biomass approximately equal to the estimated amount of dry biomass intake of current cattle population then: ASSUME model representation of P cycling is sufficient e.g. this transfer is reasonable in model Purchased Feed CROPS COW MANURE Milk SOILS Exit soil profile Fertilizer Farm Boundary

Conclusions and Future Work Refine P mass balance in absence of model for the basin Refine climate inputs Check mass balance of removed biomass versus estimated cattle consumption Possibly refine harvest amounts if necessary Redo the analysis after final model updates

QUESTIONS? The END

SWAT Spatial Input Data STATSGO Soil Coverage Land Use Classification A DEM (Digital Elevation Map) is also required