http://atlurbanist.tumblr.com/post/48357329068/news-of-the-insane-symposium-treats-sprawl-as-a-good

http://abcnews.go.com/blogs/technology/2012/09/michigan-and-bangalore-global-warming-goes-global/ http://ga.water.usgs.gov/edu/irdrip.html

Mapping Groundwater Recharge Rates Under Multiple Future Climate

Scenarios in Southwest Michigan

Mapping Groundwater Recharge Rates Under Multiple Future Climate

Scenarios in Southwest Michigan

http://mi.water.usgs.gov/reports/images/cover_med01_4227.jpg

Glenn O’NeilInstitute of Water Research – Michigan State University

2015 International SWAT ConferencePurdue University

10/16/2015

1

Mapping Groundwater Recharge

Introduction Methods Results

Introduction

Conclusions

2

Primary Objective:

Model the water table of Kalamazoo County, Michigan under various future scenarios of climate change, urbanization, and expanded agricultural production.

Sub Objective (primary for this presentation):

Use SWAT to simulate and map (at field scales) groundwater recharge under future scenarios of climate change.

Expectations:

3

Evapotranspiration will increase due to higher temperatures.

Subsequent decrease in groundwater recharge.

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

- Kalamazoo County, Michigan

- Existing MODFLOW model by USGS(Luukkonen et al. 2004)

- 40% agriculture, 21% forest, 20% urban

- well draining soils

- moderate topographic relief

- precipitation 36 in./yr.

- population 254,000, growing

- average annual water use- ag: 26 MGD- industry: 5 MGD- City of Kalamazoo: 19 MGD- City of Portage: 6 MGD

Primary Study Site

4

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

Intersecting watersheds of MODFLOW model boundary.

SWAT Study Site

5

KALAMAZOOCOUNTY

Kalamazoo River Watershed

Headwaters of the St. Joseph River

Watershed

Headwaters of the Thornapple River Watershed

Paw Paw River Watershed

Upper Dowagiac River Watershed

MODFLOW Model

Boundary

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Inputs

6

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Data Source

Land cover Cropland Data Layer (2009-2013)

Soils SSURGO

Topography USGS 10-meter DEM

Irrigation Michigan DEQ, well logs

Consumptive water use Michigan DEQ

Point sources Michigan DEQ, US EPA

Dams USACE National Inventory of Dams

Weather Maurer et al., 2002.

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Inputs

7

- Observed weather data interpolated to grid points

- 1/8 deg. resolution.

- Daily precip, min. temp., max temp.

- 1940 - 2010

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

HRU Mapping

8

- HRU thresholds: Land cover – 3%, Soil – 3%, Slope – 3%.- Back-mapped to raster format.- Resulted in 256 sub-basins, 76,281 HRUs.

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

HRU Mapping

9

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

- HRU thresholds: Land cover – 3%, Soil – 3%, Slope – 3%.- Back-mapped to raster format.- Resulted in 256 sub-basins, 76,281 HRUs.

SWAT Calibration

10

To best simulate ground water recharge, SWAT was calibrated to baseflow conditions.

Legend

# USGS Gages

Kalamazoo County

SWAT Model Watershed

04097540

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Calibration

11

USGS baseflow separation program identified days where 75% of flow was from ground water.

Observed Flow

Simulated Flow

Kalamazoo River – USGS Gage 04108660

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Calibration

12

USGS baseflow separation program identified days where 75% of flow was from ground water.

Observed Flow

Simulated Flow

Kalamazoo River – USGS Gage 04108660

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

Climate Change Simulation

13

11 different models, each under various CMIP-3 scenarios, downscaled to Maurer grid points by Hayhoe, et al. (2013).

Daily values for precipitation and temperature through 2100.

Climate Models

1. CCSM2. CGCM3-T473. CGCM3-T634. CNRM5. ECHAM56. ECHO7. GFDL-2-08. GFDL-2-19. HADCM310. HADGEM11. PCM

Climate Scenarios

1. A1FI – rapid population growth, levels off mid-century, heavy fossil fuel use

2. A1B – rapid population growth, levels off mid-century, balanced fossil fuel use

3. A2 – continuous population growth, regional economic growth

4. B1 – population growth levels off, efficient energy technologies adopted

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

Climate Change Simulation

14

Run SWAT in 10-year increments, for each decade:

- adjust .wgn files

- adjust CO2 concentrations (ppm) from IPCC*

- A1Fi: 420 (2020) – 970 (2100)- A1B: 403 (2020) – 717 (2100)- A2: 417 (2020) – 856 (2100)- B1: 412 (2020) – 549 (2100)

- run the model

- grab the outputs

* http://www.ipcc-data.org/ancilliary/tar-isam.txt

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Calibration

15

- Calibrated SWAT models from 2000-2005

- Validated from 2006-2010

- Moriasi et al. (2007) provided guidance on flow metrics

Nash-Sutcliffe Efficiency Percent Bias

Very Good 0.75 - 1.0 < 10%

Good 0.65-0.75 10 – 15%

Satisfactory 0.5 – 0.65 15 – 25%

Unsatisfactory < 0.5 > 25%

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Calibration

16

- Model performance for monthly baseflow

NSE % Bias

Very Good 0.75 - 1.0 < 10%

Good 0.65-0.75 10 – 15%

Satisfactory 0.5 – 0.65 15 – 25%

Unsatisfactory < 0.5 > 25%

SWAT Model / Gage Calibration NSECalibration % Bias

(negative value = more simulated baseflow)

Validation NSEValidation %Bias

(negative value = more simulated baseflow)

04108660 (Kalamazoo) .69 -1% 0.55 10%

04108600 (Kalamazoo) 0.02 -20% 0.89 -36%

04106000 (Kalamazoo) 0.66 -4% 0.69 -8

04105000 (Kalamazoo) 0.64 1% 0.48 <1%

04103500 (Kalamazoo) 0.86 -4% 0.58 8

04097500 (St. Joe) 0.66 -14% 0.55 3%

04097540 (St. Joe) 0.83 -16% 0.18 -13%

04096515 (St. Joe) 0.99 -50% 0.99 -65%

04096405 (St. Joe) 0.97 -6% 0.94 7%

Paw Paw 0.53 7% 0.50 -9%

Thornapple 0.61 1% 0.80 8%

Upper Dowagiac 0.49 3% 0.50 -6%

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Calibration

17

NSE % Bias

Very Good 0.75 - 1.0 < 10%

Good 0.65-0.75 10 – 15%

Satisfactory 0.5 – 0.65 15 – 25%

Unsatisfactory < 0.5 > 25%Map of Nash Sutcliffe Efficiency Values

- Model performance for monthly baseflow

1.7 cms55.2 cms

1.2 cms33.9 cms

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Calibration

18

- Did not just rely on flow.

- Compared SWAT outputs to:

- Reported crop yields in NASS- USGS estimates of ET- USGS estimates of baseflow fraction- Reported irrigation rates to M-DEQ and M-DARD- Estimates of ground-water recharge from M-DEQ/USGS/RSGIS/IWR

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

19

- SWAT Outputs for 04106000 (Kalamazoo)

800

850

900

950

1000

1050

mm

decades

Average annual precip

a1fi

a1b

a2

b1

average

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

20

- SWAT Outputs for 04106000 (Kalamazoo)

8

9

10

11

12

13

14

15

16

De

gre

es

C

decades

Average annual temperature

a1fi

a1b

a2

b1

average

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

21

- SWAT Outputs for 04106000 (Kalamazoo)

0

50

100

150

200

250

300

350

400

450

mm

/yr

decades

Groundwater recharge

a1fii

a1b

a2

b1

average

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

22

- SWAT Outputs for 04106000 (Kalamazoo)

400

450

500

550

600

650

700

750

mm

/yr

decades

Evapotranspiration

a1fi

a1b

a2

b1

average

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

23

- SWAT Outputs for 04106000 (Kalamazoo)

0

50

100

150

200

250

300

350

400

450

mm

/yr

decades

Groundwater contribution to stream flow

a1fi

a1b

a2

b1

average

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

24

- SWAT Outputs for 04106000 (Kalamazoo)

0

5

10

15

20

25

30

35

mm

/yr

decades

Runoff water contribution to stream flow

a1fi

a1b

a2

b1

average

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

25

- SWAT Outputs for 04106000 (Kalamazoo)

0

20

40

60

80

100

120

140

160

180

mm

/yr

decades

Irrigation (total)

a1fi

a1b

a2

b1

average

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

26

- SWAT Outputs for 04106000 (Kalamazoo)

6000

6200

6400

6600

6800

7000

7200

7400

7600

kg/h

a

decades

Corn yield

a1fi

a1b

a2

b1

average

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

27

- Changes in recharge (120mm, 60%) due to two primary sources:

- Increased precip over the century (100mm, 10%)- Decreased ET over the century (50 mm, 8%)

- Lower ET caused by higher CO2, plants transpire less.- from the SWAT documentation:

“As carbon dioxide levels increase, plant productivity increases and plant water requirements go down.”“Morrison (1987) found that at CO2 concentrations between 330 and 660 ppmv, a doubling of CO2 concentration resulted in a 40% reduction in leaf conductance.”

- CO2 is at 970 ppm by 2100 for A1Fi, 549 ppm for B1.

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

28

- From the SWAT Theoretical Documentation (p. 130):

= plant canopy resistance (s m-1)

= minimum resistance of a single leaf (s m-1)

= leaf area index

0

50

100

150

200

250

300

30

0

33

0

36

0

39

0

42

0

45

0

48

0

51

0

54

0

57

0

60

0

63

0

66

0

69

0

72

0

75

0

78

0

81

0

84

0

87

0

90

0

93

0

96

0

99

0

Can

op

y re

sist

ance

(s

m-1

)

CO2 (ppm)

Canopy Resistance

corn

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

29

- Holding CO2 constant in 04097540 kept recharge flat while ET rose slightly.

0

50

100

150

200

250

300

350

mm

/yr

decades

Groundwater recharge (a1fi)

ccsm-a1fi

gfdl_2-1-a1fi

hadcm3-a1fi

pcm-a1fi

average

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

30

- Holding CO2 constant in 04097540 kept recharge flat while ET rose slightly.

400

450

500

550

600

650

700

750

800

850

900

mm

/yr

decades

Evapotranspiration (a1fi)

ccsm-a1fi

gfdl_2-1-a1fi

hadcm3-a1fi

pcm-a1fi

average

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

31

- Spatial differences in HRU outputs.

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

32

- Spatial differences in HRU outputs.

Baseline recharge

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

33

- Spatial differences in HRU outputs.

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

34

- Spatial differences in HRU outputs.

Recharge change, 2010 - 2090

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

SWAT Future Simulations

35

0

100

200

300

400

500

600

2010-2019 2020-2029 2030-2039 2040-2049 2050-2059 2060-2069 2070-2079 2080-2089 2090-2099

mm

/yr

decades

Average annual recharge by land cover (a1fi)

CORN

FRSD

PAST

URLD

URML

WATR

WETF

- Corn areas in 04106000 recharged most, urban least.- Pasture started even with forest, surpassed it around 2050.

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

Limitations

36

- Land cover did not change in future simulations.

- Solar radiation and relative humidity did not change.

- Growing season parameters did not change (e.g. no double-cropping).

- Did not calibrate for nutrients, may affect crop-growth.

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

Conclusions

37

- Groundwater recharge generally decreased in future climate scenarios, except for the b1 scenario.

- ET generally increased, except for the b1 scenario, due to increased CO2 levels.

- Crop yields were flat.

- Spatially, largest increases in recharge through 2100 are in forested and pasture areas.

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

Next Steps

38

- Feed the recharge maps into MODFLOW to produce steady-state head for each decade within each climate scenario.

- Run another batch of simulations for increased urbanization (more imperviousness around urban centers, more consumptive water use), with current climate conditions.

- Run another batch of simulations for expanded agriculture (marginal lands converted to corn-soy rotations, more lands implementing irrigation), with current climate conditions.

- Run a final batch of simulations combining the increased urbanization, agricultural expansion, and changing climate.

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions

References

39

Hayhoe, K. 2013. Development and dissemination of a high-resolution national climate change dataset. Final Report for United States Geological Survey, USGS G10AC00248 (Accessed online 10-08-2014 at: https://nccwsc. usgs. gov/displayproject/5050cc22e4b0be20bb30eacc/4f833ee9e4b0e84f608680df).http://cida.usgs.gov/thredds/catalog.html?dataset=dcp

Luukkonen, Carol L., Stephen P. Blumer, T.L. Weaver, and Julie Jean. 2004. “Simulation of the Ground-Water-Flow System in the Kalamazoo County Area, Michigan.” 2004-5054. USGS Scientific Investigations Report. Reston, Virginia: U.S. Geological Survey. http://pubs.usgs.gov/sir/2004/5054/ .

Maurer, E. P., A. W. Wood, J. C. Adam, D. P. Lettenmaier, and B. Nijssen. 2002. “A Long-Term Hydrologically Based Dataset of Land Surface Fluxes and States for the Conterminous United States*.” Journal of Climate 15 (22): 3237–51. doi:10.1175/1520-0442(2002)015<3237:ALTHBD>2.0.CO;2.http://cida.usgs.gov/thredds/catalog.html?dataset=cida.usgs.gov/new_gmo

Ground-water Sustainability (Update)

Introduction Hypotheses Methods Results Next Steps

Mapping Groundwater Recharge

Introduction Methods Results Conclusions