hydrologic research center [email protected] cpasw march 23, 2006 incorporating climate...

Hydrologic Research Center http://www.hrc-lab.org [email protected]

CPASW March 23, 2006

Incorporating Climate Variability Uncertainty in Water Resources Planning

for the Upper Santa Cruz River.

Eylon Shamir

CPASW March 23, 2006

Incorporating Climate Variability Uncertainty in Water Resources Planning

for the Upper Santa Cruz River.

Eylon Shamir

Hydrologic Research Center

12780 High Bluff Drive suite 250San Diego, CA 92130

Tel: (858) 794 2726

www.hrc-web.org


ProjectProject

Researchers: Konstantine Georgakakos, HRC Director Eylon Shamir, HRC Nicholas Graham, HRC Jianzhong Wang, HRC David Meko, The Tree-Ring Research Laboratory, The University of

Arizona

In cooperation with Arizona Department of Water Resources: Frank Corkhill, Alejandro Barcenas, Frank Putman, Gretchen Erwin and

Keith Nelson.

Sponsored by,

Arizona Department of Water Resources Contract No. 2005-2568


Mexico

U.S

Santa C

ruz River

Sa

nta

Cru

z Riv

er

Precipitation and Streamflow gaugesNogales

Tucson

Mexico

Santa Cruz Headwater


The Study ObjectivesThe Study Objectives

Develop a modeling system that produces likely future streamflow scenarios at the Nogales USGS Gauge site.

Integrate the future streamflow scenarios with a groundwater model

Evaluate the future streamflow – groundwater response in various schemes of water consumptions.


Zoom into the study areaZoom into the study area

Data

Analysis Area

Area of Interest


Landscape Landscape


Trends in summer (Jul-Aug) flowTrends in summer (Jul-Aug) flow

Rain

Flow

Rain

Flow

1910 1920 1930 1940 1950 1960 1970 1980 1990 20000

5

10

15

YearsC

ount

of

Eve

nts

NogalesLaveenTucsonLochiel

1910 1920 1930 1940 1950 1960 1970 1980 1990 20000

0.2

0.4

0.6

0.8

1

Years

Nor

mal

ize

d to

tal s

umm

er

flow

Nu

mb

er

of

Eve

nts

No

rmal

ized

To

tal

Su

mm

er F

low

Pool and Coes (1999) found similar trend in Charleston gauge –San Pedro


Variability in Monthly FlowVariability in Monthly Flow

0 200 400 600 800 10000

2000

4000Monthly volume (106 ft3)

0 200 400 600 800 10000

10

20Box-Cox transformed

0 200 400 600 800 1000-5

0

5Seasonal standardization of the Box-Cox transformed

Months (Jan 1936 -Sep 2003)

0 200 400 600 800 1000-2

-1

0

1

2

Mea

n

Mean of the moving average 24 month

0 200 400 600 800 10000

0.5

1

1.5

2

S.D

Linear fit of the S.D of the moving average 24 month

• Change in monthly flow variability since the 1970s

Average monthly flow as a function of time

Variability of monthly flow as a function of time

1936 Years 2003

0 1000 2000 30000

200

400

600

800(a) Monthly flow

Cou

nts

-4 -2 0 2 40

20

40

60

80(b)Transformed and standardized

-4 -2 0 2 40

20

40

60

80(c) Normal distribution

Cou

nts

-4 -2 0 2 40

20

40

60

80 (d) Conditional normal distribution

0 1000 2000 30000

200

400

600

800(a) Monthly flow

Cou

nts

-4 -2 0 2 40

20

40

60

80(b)Transformed and standardized

-4 -2 0 2 40

20

40

60

80(c) Normal distribution

Cou

nts

-4 -2 0 2 40

20

40

60

80 (d) Conditional normal distribution

Histograms


Nogales precipitation Vs. climatic indicesNogales precipitation Vs. climatic indices

Correlations with Nogales precipitation (1915-2000).NINO3 PDO ARIZ.

DIV. 7

WINTER 0.53 0.27 0.94

DRY 0.11 0.22 0.70

SUMMER -0.06 0.09 0.53

OCTOBER -0.03 -0.09 0.87

• Only the winter flow in Nogales is correlated with El–Niño

Pre

cip

ita

tio

n (

inc

he

s)

Nino SST Anomaly (°C)

Nogales Annual Precipitation Vs. NINO3 SST Anomaly (1915-2000)

R=0.53

Climate Divisions


Simulation of Precipitation Vs. Streamflow Simulation of Precipitation Vs. Streamflow

Precipitation (Pros) Better linked to climatic forcing and global circulation Less affected by geomorphological changes and human

activity in regional scale. Independent of the basin antecedent condition

Precipitation (Cons) Point measurement rather than areal measurement that

contributes to the flow. Requires a model to transform into streamflow


0 50 100 150 200 250

0.1

0.2

Inch

es

Winter 1969 (Hours)

0 50 100 150 200 250

0.5

1

1.5

Inch

es

Summer 1969 (Hours)

Winter Hourly Precipitation (Dec 68- Feb 69)

Summer Hourly Precipitation (Jul 69- Aug 69)

Precipitation events

Precipitation clusters

Clusters inter-arrival time

0 50 100 150 200 250

0.1

0.2

Inch

es

Winter 1969 (Hours)

0 50 100 150 200 250

0.5

1

1.5

Inch

es

Summer 1969 (Hours)

Stochastic Precipitation model components

Winter


Daily streamflow transformation model for the basin outlet

Groundwater storage changes in the downstream aquifers

Long term risk assessment for a variety of management schemes

Stochastic hourly precipitation over the basin drainage area

Daily streamflow transformation model for the basin outlet

Groundwater storage changes in the downstream aquifers

Long term risk assessment for a variety of management schemes

Stochastic hourly precipitation over the basin drainage area

The Modeling Scheme The Modeling Scheme


0

300

600

900

1 6 11 16 21

0

300

600

900

1 6 11 16 21

Days

Dis

char

ge (

cfs)

Winter (1/4/95)

Summer (8/27/99)

Nogales Daily Precipitation 1980

0

5

10

15

20

25

30

35

40

45

50

0 366

1980 (Days)

Pre

cip

itat

ion

(m

m)

Winter

Summer

Summer and winter properties

Model of the watershed

• The model should maintain the special characteristics of the seasons


Total Seasonal Flow Total Seasonal Flow

Seasonal division:Winter: November-MarchSpring: April–JuneSummer: July-SeptemberFall: October

0 20 40 60 800

1

2

3x 10

4 Summer

Tot

al s

easo

n (c

ubic

3 )

0 20 40 60 800

2

4

6x 10

4 Winter

0 20 40 60 800

1

2

3x 10

4 Fall

Tot

al s

easo

n (

feet

3 )

Years0 20 40 60 80

0

1000

2000

3000Spring

Years

To

tal

Sea

so

nal

Flo

w (

ft3)


0 50 1000

20

40

60(a) Summer

SimulatedObserved

0 50 1000

20

40

60(d) Spring

0 50 1000

20

40

60(b) Winter

0 50 1000

20

40

60(b) Fall

0 20 40 60 80 100500

1000

1500

2000

2500

3000

3500

Percentiles

Bo

x-C

ox

Tra

nsf

orm

ed

(ft3

)

Evaluation of the simulated streamflow

0 50 1000

1000

2000

3000(a) Summer

ObservedSimulated

0 50 1000

1000

2000

3000(b) Winter

0 50 1000

1000

2000

3000(c) Fall

Percentiles

T

rans

form

ed

Sea

son

al v

olu

mes

0 50 1000

500

1000

1500(d) Spring

Annual

Daily Seasonal

Percentiles

Tra

ns

form

ed

Bo

x-C

ox

Da

ily

Flo

w


Precipitation estimates from tree-ringsPrecipitation estimates from tree-rings

Current record

Reconstructed by, David Meko, The Tree-Ring Research Laboratory, The University of Arizona


Winter flow categories

Precipitations quartiles of the tree rings

1st 2nd 3rd 4th

Wet 0 0 0 0.625

Medium 0.5 0.69 0.75 0.375

Dry 0.5 0.31 0.25 0

Streamflow scenarios forced by the tree-ring reconstructed precipitations


Hig

hway

82

ArizonaMexico

International Wastewater Treatment Plant

Precipitation gauge

Precipitation gauge

Streamflowgauge

Po

trero creek

Santa Cruz River

Well 1

Well 2

Well 3

Well 4Well 5

Well 6Well 7

Well 8

Well 9

Wells# Microbasins

Buena Vista

Guevavi

Highway 82

Kino Springs

Elevation (m)

1000 - 1100

1100 - 1200

1200 - 1300

1300 - 1400

1400 - 1500

1500 - 2300

No Data

Elevation (m)

Groundwater Microbasins


0 10 20 30 40 50 600

0.51

Buena Vista

0 10 20 30 40 50 600

0.51

Kino Springs

0 10 20 30 40 50 600

0.51

Hy 82

Re

lativ

e F

illin

g

0 10 20 30 40 50 600

0.51

Guevavi

Months10/97 9/02

Red –AZDWR MODFLOW MODELBlue- HRC Simplified model

Groundwater Model Development

Q1

I1 Q2 = Q1-I1Santa Cruz River

ET1

P1

Buena Vista(2740 ac/ft)

Kino Springs(4020 ac/ft)

Highway 82(5910 ac/ft)

Guevavi7950 (ac/ft)

Q1

I1 Q2 = Q1-I1Santa Cruz River

ET1ET1

P1

Buena Vista(2740 ac/ft)

Kino Springs(4020 ac/ft)

Highway 82(5910 ac/ft)

Guevavi7950 (ac/ft)


0 500 1000 1500 20000

1020

Buena Vista

0 500 1000 1500 20000

50100

Kino Springs

0 500 1000 1500 20000

50100

Hy 82

Dep

th t

o W

ate

r (f

t)

0 500 1000 1500 20000

50Guevavi

10/97 9/02Days

Model Comparison with index-wells

Point to area


How can the model output be used?How can the model output be used?

Time

Flow

Ensemble of possible scenarios

Threshold

Water consumption

Well

Groundwater Storage

Chance to drop bellow the t-hold

Aquifer

Risk Analysis


Various water consumption scenarios


Ensemble with 100 realizations

Ensemble with 1000 realizations

Consecutive monthly stress

Consecutive Months Below 50%

Consecutive Months Below 50%


0 20 400

0.5

1Buena Vista

0 20 400

0.5

1Kino Springs

0 20 400

0.5

1Hy 82

Pumpage (% of reservoir capacity)

F

ract

ion

of

time

bel

ow

thr

esh

old

0 20 400

0.5

1Guevavi

Risk Assessment Using Tree Ring Winter Precipitation Estimate


Future ChallengesFuture Challenges

The use of risk analysis as exemplified in this work in collaboration with regional officials and agencies to establish policy regarding regional development.

Incorporation of climate change scenarios to possibly improve the generation of future streamflow ensembles.

Application in other semi-arid or arid regions


http://www.hrc-lab.org/projects/dsp_projectSubPage.php?subpage=santacruz

Project Report:


Precipitation EvaluationPrecipitation Evaluation

Medium Summer

Red –Observed Blue -Simulated


Nogales GaugeNogales Gauge

Nogales gauge hydrograph

-2000

0

2000

4000

6000

8000

10000

12000

14000

19

35

19

37

19

39

19

41

19

43

19

45

19

47

19

48

19

50

19

52

19

54

19

56

19

58

19

60

19

62

19

64

19

65

19

67

19

69

19

71

19

73

19

75

19

77

19

79

19

81

19

82

19

84

19

86

19

88

19

90

19

92

19

94

19

96

19

97

19

99

20

01

20

03

Days

Me

an

da

ily

flo

w (

cfs

)

0

500

1000

1500

2000

2500

1 4 6 9 12 3 6 9 12 3 6 9 12 3 6 9 12 3 6 9 12

Month (1998-2002)

Dis

char

ge

(cfs

)


0 2 4 6 8 100

0.5

1

1.5

2

2.5

3

3.5

4

4.5

Days

CV

SummerWinterFallSpring

Minimum Cluster Inter-Arrival Time

Co

effi

cien

t o

f V

aria

tio

n o

f I

nte

r A

rriv

al P

erio

d

Restrepo-Posada and Eagleson (1982)


0 20 400

2000

4000

cfs

days0 20 40

0

20

40Jan 1979 Total Precipitation

0 20 400

100

200

cfs

days5 10 15 20

0

20


Cou

nt

5 10 15 200

20


Total Precipitation (cm)0 20 40

0

20

40

cfs

days

Regional simulation using mm5 atmospheric model 6X6 km, 20 second (output at 1 hour) for January 1979, 1991, and 1992 Lateral Boundary layers are from the NCEP ETA re-analysis data 32X32 km 3 hour

Precipitation distribution from regional atmospheric modeling Precipitation distribution from regional atmospheric modeling

January 16 1979January 16 1979

Wind Speed and Direction


Precipitation areal distributionPrecipitation areal distribution

How is precipitation distributed over the area?

With the lack of dense raingauges, we used: Regional atmospheric model with high spatial resolution

Analysis was done for 6 historical winter storms


Areal distribution of precipitation for 6 winter stormsfrom regional atmospheric model


Stochastic Hourly Precipitation Model


B

xA

x eBf)(

1

, A ≤ x, and B > 0

Exponential Distribution



Exponential Distribution cont.Exponential Distribution cont.Table B-2: Parameter values of the exponential distributions that are used to simulate the hourly precipitation at Nogales. In parenthesis are the fitted parameters that are used to match the flow. Cluster inter arrival

period Duration of

cluster Hourly precipitation magnitude

A B A B A B Winter: Wet 0.06 0.17 0.02 0.26 0.02 0.1 (0.5)

Medium 0.11 0.4 -0.05 0.1 0.008 0.14

Dry 0.12 0.3 -0.04 0.1 0.08 0.12

Summer: Wet 0.06 0.4 (0.1) 0.02 0.17 -0.04 0.13 (0.5)

Medium 0.04 0.5 (0.15) -0.01 0.3 (0.2) -0.03 0.1

Dry 0.02 0.45 (0.4) -0.015 0.2 -0.04 0.15 (0.1)

Hourly precipitation chance

Magnitude

Fall -0.03 0.2 (0.05) -.009 0.04 (0.015)

Spring 0.04 0.24 -0.01 0.087 (0.01)

0 5 10 15 20 25 300

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

x

f(x)

B=0.1

B=0.4



0

300

600

900

1 6 11 16 21

0

300

600

900

1 6 11 16 21

Days

Dis

char

ge (

cfs)

Winter (1/4/95)

Summer (8/27/99)

Processes-based Conceptual Model

Hourly precipitation to mean daily flow


Seasonal Daily flow of the three moments

simulations are from 100 realizations 100 year each. (mean and the standard deviation of the moments from the 100 realizations).


Ensemble of 100 realizations using the tree ring reconstruction of precipitation

Ensemble of 100 realizations using the tree ring reconstruction of precipitation

hydrologic research center [email protected] cpasw march 23, 2006 incorporating climate...

Documents