Holocene Flood Frequency in New England:Large, Episodic Events in the Sediment Record

Adam ParrisAdam ParrisUniversity of VermontUniversity of Vermont

Master’s Thesis DefenseMaster’s Thesis DefenseApril 18, 2003April 18, 2003

WHY?WHY?

2000

2600

5900

68008200

9400

50 cm

Cal

enda

r Yea

rs B

P

MAGLOI

Brown et al.,Geology, 2000

New EnglandClimate

0 2000 4000 6000 8000 10000 12000

Cool, moist

Warm, dry

Cool, moist

Storms

Fans

Avg.

Jennings et al.,GSA Bulletin, 2003

Noren et al., Nature, 2002

0 2 4 6 8 10 12

Calendar ky BP

Questions Does particle size anlaysis of lake sediment cores

offer a higher resolution record of storminess?

Are there periods and/or cycles of increased storminess in NH & ME, similar to VT & NY?

Is the regional storm record in NH & ME similar to other Atlantic records?

●Climatic causes/controls?

Reasoner Method

Driver

Head

Barrel

Piston

6m

LakeLocations

DEEPDEEP STEEPSTEEP

Lake NameSurface

Area (km2)

Max Depth

(m)Basin

Relief (m)Worthley Pond

(ME) 1.43 15 344Crystal Lake

(NH) 0.4 18 353South Pond

(NH) 0.7 27.9 427Ogontz Lake

(NH) 0.303 22 408Stinson Lake

(NH) 1.4 22 655Sandy Pond

(NH) 0.11 12 226

Analytical MethodsAnalytical Methods●● Visual Stratigraphic LogVisual Stratigraphic Log● ● Magnetic Susceptibility (MAG)Magnetic Susceptibility (MAG)● ● Loss on ignition (LOI)Loss on ignition (LOI)● ● AMS- Radiocarbon Analysis (AMS- Radiocarbon Analysis (1414C)C)

● ● Particle Size AnalysisParticle Size Analysis ●●COMPREHENSIVE PREPCOMPREHENSIVE PREP ●●HIGH RESOLUTIONHIGH RESOLUTION ● ●REGIONAL COVERAGEREGIONAL COVERAGE

Med PS (µm)

% LOIMAG (SI) VisualLog

20

60

100

140

180

Dep

th (c

m)

15 25 350 6 12 1 10 1001 10 100 1000

Mean PS (µm)

MedPS (µm)

PS%

Vol

ume

1 10 100

20

60

100

140

180

Dep

th (c

m)

1 10 100 1000

MeanPS (µm)

10 1000 10 1000 10 100010 100 100 5 5525-5

100

200

Dep

th (c

m)

Mean(µm)

Median(µm)

Mode(µm)

StDev(µm)

Kurtosis Skewness

MedPS (µm)

PS%

Vol

ume

1 10 100

20

60

100

140

180

Dep

th (c

m)

1 10 100 1000

MeanPS (µm)

Size Frequency Distributions

0

0.5

1

1.5

2

2.5

3

0.1 1 10 100 1000

Particle Size (µm)

Vol

ume

%

End Member Modeling:Unraveling the size distribution

60 % 20 % 20 %

0.0

0.2

0.4

0.6

0.8

1.0

1 2 3 4 5 6 7 80.0

0.2

0.4

0.6

0.8

1.0

0.1 1 10 100 1000

Particle Size (µm)Coe

ffici

ent o

f Det

erm

inat

ion

(r2 )

Number of end members

Med

ian

r2

Estimation of End Members

Particle Size (µm)

Vol

ume

%

0.0

3.0

6.0

9.0

12.0

0.1 1 10 100 1000

EM5:EM5:Mode= 6Mode= 6µmµm

EM4:EM4:Mode= 19Mode= 19µmµm

EM3:EM3:Mode= 58Mode= 58µmµm

EM2:EM2:Mode= 121Mode= 121µmµm

EM1:EM1:Mode= 373Mode= 373µmµm

End Members

Proportion of the End MembersD

epth

(cm

)

100

200

300

0.0 1.00.5 0.0 1.00.5 0.0 1.00.5 0.0 1.00.5

EM1EM1 + + EM2EM2 EM3EM3 EM4EM4 EM5EM5

Particle Size (µm)

Volu

me

%

0.0

3.0

6.0

9.0

12.0

0.1 1 10 100 1000

EM5:EM5:Mode= 6Mode= 6µmµmEM4:EM4:Mode= 19Mode= 19µmµmEM3:EM3:Mode= 58Mode= 58µmµmEM2:EM2:Mode= 121Mode= 121µmµmEM1:EM1:Mode= 373Mode= 373µmµm

• 80 dates• John Southon, Tom Guilderson Lawrence Livermore National Laboratory

Stinson Lake Core 2 Age Model

0100200300400500600

0 2000 4000 6000 8000 10000 12000Age (yr)

Dep

th in

Cor

e (c

m)

Calibrated Age

Radiocarbon Age

Time Series Filter

● SSA construction of raw data series

● Remove linear trend of data

● Peaks > 1Peaks > 1σσ from from reconstruction = reconstruction = SIGNIFICANT EVENT LAYERSIGNIFICANT EVENT LAYER

200

225

250

275

300

PS

25

65

105

145

185

Dep

th (c

m)

10 40 0 1 0 200 1 100

Mean PS(µm)

Median PS(µm)

%LOI EM1 + EM2

 Mean Median Min Max

LOI 1σ 31 28 15 70

LOI 2 σ 8 4 2 32

Mean PS 1 σ 39 33 11 81

Mean PS 2 σ 20 18 7 45

Median PS 1 σ 37 32 12 88

Median PS 2 σ 18 15 6 52

Coarse EMs 1 σ 29 22 14 57

Coarse EMs 2 σ 15 12 7 31

Event Detection Comparison

Removeeventlayers

Compresscore

VLVLVL50

100

150

200

250

300

VL50

100

150

200

250

300

Events

1400

3000

5200

Dates

14002100, 2150

Model Ages

3000, 31203400, 3790

4800, 49505200

0 2 4 6 8 10 12 14Cal. Age (kyr BP)

Terrigenous LayerEnd of recordBeginning of record

Storm Dates

0 2 4 6 8 10 12

Cal Age (kyr BP)

Storm Frequency0.80.8 1.41.4 2.12.1 3.03.0 3.93.9 6.86.8 11.511.5

0

0.05

0.1

0.15

0.2

0.25

0.3

0.000 1.000 2.000 3.000 4.000 5.000

Raw Spectrum

Harmonic

Median

90% CI

95% CI

99% CI

Frequency (cycles/kyr)

Storm Frequency- EMM

5600 yrs600 yrs

311 yrs

Climate in Northeastern USA

StormFrequency

Storms

Floods

0 2 4 6 8 10 12Calendar kyr BP

Hurricane Connection

• Increased storminess in New England = Increased hurricane frequency in Atlantic and Gulf Coast

• Increased storminess in New England = Increased flooding in Upper Mississippi Valley

• Modern record of hurricane-related precipitation and flooding

Discussion

• Differing landscape response to precipitation

• Storm size– More frequent events– Smaller cycles– Radiocarbon uncertainties

Discussion

• Geographic location of study area

• Proposed mechanisms for hurricane-related storms in New England

Conclusions

• Grain size is an effective physical parameter which detects more frequent events

• End Member Modeling is a valuable tool for revealing processes reflected in grain size distributions

• Different patterns of storminess in NH & ME than VT & NY

• Patterns of storminess in NH & ME connected to hurricane related storms

Acknowledgements

Paul BiermanPeter RyanCharlotte MehrtensAngie ConlanSarah BrownKristen BenchleyJoseph HaunSarah Stopper

Andy BosleyStephen WrightKaren JenningsJohn SouthonTom GuildersonEric SteigCraig Kochel

Without the financial support from NSFand the friendship, support, and collaboration of these people, this project would not be possible.

Anders NorenAndrea LiniAndi LordMaarten PrinsWayne ParrisKathy ParrisKrista ParrisLaura Mallard