Assessing the hydrologic and water quality impacts of climate change in small agricultural basins of the Upper Chesapeake Bay watershed

Anthony Buda USDA Agricultural Research Service

Monday, March 7, 2016 Westin Hotel Annapolis, MD The Development of Climate Projections for Use 1:40 to 2:00 PM in Chesapeake Bay Program Assessments Today’s presentation Upper Chesapeake Bay climate assessment Upper Chesapeake LTAR A brief overview of the four basins comprising the Upper Chesapeake LTAR location in Pennsylvania.

Climate change projections Future climate predictions for the Mahantango Creek basin using statistically downscaled data.

Modeling hydrology and water quality An approach to predicting the effects of climate change on hydrology and water quality with the SWAT model. USDA’s LTAR Network serving as a sentinel to changes in climate and hydrology

Northern Great Plains Heartland Northern Crescent

R.J. Cook Agronomy Farm Northern Plains Great Basin Kellogg Upper Mississippi Upper Chesapeake Bay Biological Station Central Plains Platte River River Basin Eastern Experimental Range High Plains Aquifer Central Mississippi Corn Belt Lower Chesapeake Bay River Basin Eastern Uplands

Southern Lower Mississippi Plains Southern Seaboard Walnut Gulch Jornada River Basin Gulf Atlantic Experimental Watershed Experimental Range Southern Coastal Plain Fruitful Rim Plains Archbold Basin and Range Biological Station

Prairie Gateway Mississippi Portal LTAR Shared Research Strategy, 2013 Upper Chesapeake Bay LTAR Four basins typifying variable physiography and farming

Anderson Creek Watershed Rockton, PA

Anderson Allegheny Creek Plateau Spring Creek Watershed Rock Springs, PA Mahantango Spring Ridge & Creek Creek Valley

Piedmont Conewago Creek Mahantango Creek Watershed Klingerstown, PA Coastal Plain

Conewago Creek Watershed Elizabethtown, PA The Mahantango Creek Watershed an ideal place to assess long-term trends in hydroclimate Precipitation (1968 to present) WE-38 Watershed (7.3 km2)

Temperature (1978 to present)

Streamflow (1968 to present)

Buda et al., 2011a, b (WRR) Significant hydroclimatic trends in WE-38 Steadily rising temperatures and a lengthening growing season are Annual 0.38 C mean temp. per decade consistent with a warming climate.

At the basin level, evapotranspiration Annual 37 mm increases represent the clearest change AET per decade in hydroclimate over the past 45 years.

More intense rain storms in the fall have led to augmented streamflow during a Tropical Storm Lee time when flows are normally low.

Snowmelt runoff events are declining in Flood of winter whereas in summer, low flow 1996 periods are expanding in duration. Climate change in the Upper Chesapeake Using downscaled data to project future climatic conditions Regional models disaggregated to 1/16

Anderson Creek Mahantango Spring Creek Creek Conewago Creek

Downscale 9 models from the Mahantango Creek watershed Coupled Model Inter- WE-38 comparison Project Phase 5 (CMIP5) for the business as usual (RCP 8.5) and stabilization (RCP 4.5) emissions pathways. Hayhoe et al., 2007 (Climate Dynamics); Stoner et al., 2013 (International J. Climatology) We obtained climate projections from nine different climate models

Climate models recommended by Katharine Hayhoe

CCSM4 HadGEM2-CC inmcm4 CSIRO-Mk3-6-0 CNRM-CM5

MIROC5 IPSL-CM5A-LR MRI-CGCM3 MPI-ESM-LR International Centre for Earth Simulation

Daily climate variables Mean temp. (C) Max temp. (C) Min temp. (C)

Solar radiation (MJ/m2) Precip. (mm) Wind speed (m/s) Mid-century temperatures will be warmer with increases of about 2C relative to 1960-1989 Hadley Centre Climate Model Version 3 (HadCM3)

Mean annual temperature 20

Mahantango C) Creek  15 10 5

Temp. ( Temp. 0 1950 2000 2050 2100

Mid-century (2015 to 2044) increase in mean annual temperature F 3.0 3.5 4.0

C 1.67 1.94 2.22 A warmer climate means more extremes daily max temperatures may approach 42C by century’s end

60

C)  50

40

30 Highest daily max Highest daily max Highest daily max 20 temp in 1975 temp in 2025 temp in 2075 35 (3)C 37 (4) C 42 (6) C

10 Annual max value of daily max temp. ( temp. of max daily value Annual max

0 1950 1975 2000 2025 2050 2075 2100 Year Mid-century will be wetter increases will range from 5 to 15% relative to 1960-1989

Hadley Centre Climate Model Version 3 (HadCM3)

Mean annual precipitation 2000 Mahantango Creek mm) . . ( 1000

0 Precip 1950 2000 2050 2100

Mid-century (2015 to 2044) increase in mean annual precipitation % 5 10 15 Daily rains of 25 mm will be more routine with 5 more such days by the year 2100

25

20

15

10

5 Days per year with rainfall > mm 25 with rainfall peryear Days

0 1950 1975 2000 2025 2050 2075 2100 Year More frequent 20-yr storms (127 mm/d) with a 3-fold increase in frequency expected by 2100

Northeast By late century, events that recurred once in 20 years may happen three to four times as often.

Future change multiplier

1 2 3 4 5 6 7

Wuebbles et al., 2014 (BAMS); US National Climate Assessment, 2014 Paradoxically, the future also could be drier Evaporative demand is likely to overwhelm inputs from rain Longer dry spells Increased risk of drought

Northeast

Change in max consecutive dry days (%) Standardized soil moisture (0-30 cm; deviations from 20th century mean) for 2090 to 2099 using the RCP 8.5 emissions scenario (Cook et al., 2015; Science Advances). More than 80% of climate Take home point: more rain is needed models suggest that successive to keep pace with rising evaporative dry days will rise by 5 to 10%. demand (Sherwood and Fu, 2014).

US National Climate Assessment, 2014; Sherwood and Fu, 2014 Implications for the Upper Chesapeake Simulating potential climate change impacts in small basins 30 ) • Emissions scenarios RCP 4.5 and 8.5 yr 25 20 • Three time frames 15 1. Early century (2015 to 2034)

emissions 10 2 2 2. Mid century (2045 to 2064)

CO 5 3. Late century (2081 to 2100)

0 (billion (billion tons/ metric 1950 1975 2000 2025 2050 2075 2100

Topo-SWAT (Easton et al., 2008)

Improved simulation of variable source area (VSA) hydrology in agricultural watersheds of the humid Northeast. Climate change and watershed hydrology How will the water balance change with climate change?

1400

1200

1000 Precipitation

800

600

400 Streamflow

200 Precipitation or streamflow depth (mm) or streamflow Precipitation

0 1950 1975 2000 2025 2050 2075 2100 Year Climate change and watershed hydrology How will climate change affect runoff generation patterns?

Using Topo-SWAT to identify variable source areas (VSAs)

Runoff (mm)

400 500 600 700 800 Mattern Watershed runoff 22,000 L

90 L well drained soil

Mattern poorly drained soil (restrictive layer) SWAT images courtesy of Tamie Veith and Amy Collick Climate change and watershed hydrology How will the frequency of floods and low flows change?

100

10 Will the flow duration curve

) shift with climate change? cms 1

0.1 Mean daily flow Mean flow ( daily

0.01

High flow Low flow 0.001 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percentage of time equaled or exceeded Climate change and water quality Effects of current climate and land use on water quality Converted Converted ConvertedForested Forested Grass Current land management basedForestedGrass on farmer surveys None GrassNone

NoneConverted PerennializationConverted ForestedCGAA Forested CGAAGrassCPSTForested buffer GrassGrass buffer CPSTNone NoneNone GRFL

ConservationGRFLCGAA Cropping CGAA CPSTDGABCover crops CPST

DGAB No till EGAB GRFL DGABEGABConservation tillage EGAC

EGAB EGAC BaselineFGAC land use EGACDGABEGABFGAC LU_WE38_currentNo change in mgmt. FGAC EGACLU_WE38_current LU_WE38_currentEGABFGAC EGACLU_WE38_current SimulateFGAC N, P, and sediment losses in runoff LU_WE38_current

Images courtesy of Tamie Veith and Amy Collick Climate change and water quality

None Relative effects of climate and land useGrass change on water quality Forested Converted Land mgmt. and BMP allocation usingLU_WE38_manureinj PA’s WIP (thru 2025)

ConvertedLU_WE38_CoverCrops Converted Farm InfrastructureForestedLU_WE38_GravelRoads Forested ConvertedGrassLU_WE38_FarmStrcBarnyard runoff Grass ForestedLU_WE38_NoTill PerennializationNone GrassLU_WE38_MinTillNone Converted Grass/forested buffer NoneConverted GrassForestedCGAAForested buffer ForestedGRFLCGAA ForestedGrassConvertedCPSTGrass buffer CGAAGrassLU_WE38_scenariosCPST NoneForested None CPSTNone LU_WE38_manureinj GrassGRFLTree planting GRFL LU_WE38_CoverCropsCGAANone ConservationCGAA Cropping LU_WE38_GravelRoadsGRFLCPST NoneDGAB LU_WE38_FarmStrc CPSTCGAADGABCover crops Grass LU_WE38_NoTillDGAB CPSTNone No till ForestedEGAB LU_WE38_MinTill GRFLGrassEGABConservation tillage ConvertedEGAC EGABGRFLForestedEGAC Manure/NutrientLU_WE38_covercropsFGAC Management GRFLEGACDGAB ConvertedFGAC LU_WE38_manureinjLU_WE38_currentManure injection LU_WE38_scenariosFGAC DGABLU_WE38_manureinjLU_WE38_current LU_WE38_NoTill

Urban StormwaterLU_WE38_currentEGABLU_WE38_CoverCropsManagement LU_WE38_MinTill EGAB EGACLU_WE38_GravelRoadsGravel road improvement EGAC FGACLU_WE38_FarmStrc BaselineGRFLFGAC land use LU_WE38_currentLU_WE38_NoTill LU_WE38_scenariosLU_WE38_currentNo change in mgmt. Images courtesy of Tamie Veith and Amy Collick LU_WE38_MinTill

GRFL LU_WE38_scenarios Acknowledgments

Ray Bryant, Peter Kleinman, Amy Collick, Gordon Folmar, Sarah Goslee, Tamie Veith, Al Rotz

Anne Stoner Katharine Hayhoe