The Hydro-Ecology of Everyday Life:
Assessing the social and environmental determinants of water use
in the Portland region
Vivek Shandas
System Science Seminar, PSU, May 2010
Urban Studies and Planning
Center for Urban Studies
Portland State University
How many Olympic size swimming
pools does it take to quench the
thirst of people living in the Portland
metropolitan region for one day?
(Hint: One Olympic Size Swimming Pool contains: 2,500 m3 or 660K Gallons)
Almost 408!
At current rates of population growth and
water use, by 2040 the region will consume the
amount of water equivalent to Crater Lake.
Water as the New Carbon?
“Water flows uphill towards money.”
–
•
Marc Reisner, Cadillac Desert, 1986
Headline news:
–
“Water Shortages: Atlanta's Cup Nearly Runneth
Out” -- The NY Times (2008)
–
“Large companies not prepared for water shortage”
– Bussiness Week (2007)
–
Water rationing is possible this summer: Light
snowpack, low reservoirs creates shortages in
California. – LA Times (2007)
–
“Shortage of water brings restrictions” – St.Petersburg
(2006)
Background – the Problem
• Population growth in the NW
• Climate destabilization and variability – supply side focus with
limited empirical evidence on demand measures
• Limited coordination between land use planning and water resource
management
– Water demand management without info on the impact of land use
patterns
– Land use planning without consideration of water demand
Residential Dwelling Density:
2005
Residential Development:
1990 - 2000
Residential Land Use
Study Objectives
• Assess the role of urban form in water use
• Investigate what factors explain patterns of
water use (demand focus)
• Examine the effects of temperature change
on water use
Conceptual Model
Coupling of Human and Environmental Systems
Structural
Environmental
Conditions
Temperature
Precipitation
Building area, zoning, pricing
scheme, urban growth, etc.
Societal
Neighborhood, demographics,
cultural, biophysical, etc.
Droughts
Individual
Preferences, habits, life
stage, financial, etc.
Total Water
Consumption
Household,
Neighborhood,
Region, State,
Country, Globe
Bull Run Watershed
• Portland's planning policies:
• No Requirement to ensure
adequate water supply
• Goal 2, of the Portland's landuse plan identifies `water' as an
infrastructure subsystem
• Bull Run Watershed
•102 square miles, Cascade
Snowmelt snowmelt
• Started in January 1895
consisted
Case Study
Portland Water Bureau
– Serves 19 water
providers of the
region
– 802,000 residents
served total, retail
service 539,200
– This study focused on
Portland’s retail
service area only
– 149,000 residential
accounts
– 30,000 commercial
and industrial
accounts
Data and Methods
• Water use (PWB)
(1) Yearly: for all taxlots (1999 – 2005)
(2) Daily: 501 SFR taxlots
• Land use (Metro)
– Structural attributes, building age, address
• Daily temperature (OSU)
• Demographic data (US Census)
– Income, education
• Arial photos (City of Portland)
– Canopy cover – parcel scale
• GIS and statistical methods
– GIS: to integrate structural and societal
characteristics
– OLS, pooled-cross section, and hierarchal
statistical models to assess relationships
Three Analysis to Address Objectives
(1) Analysis of land use building area with annual water
consumption
•
Provides evidence about the role of land use planning on water
demand management
(2) Analysis of residential and neighborhood development
patterns and annual water consumption
•
Provides evidence about the role of urban form in water demand
management
(3) Analysis of historic and projected daily maximum
temperatures with water use
•
Provides evidence about the impact of climate change on water
demand management
Synthesis of Residential Water Use Data
Results: Pooled Cross Section
• Zoned and Building area
– Overall explanatory power (R2) is improved
– Zoned area and building are coefficients vary
• MFR and SFR have similar building area water use
coefficients
Land Use Type
Coefficients
Zoned
Building Area
Com_Ind_Acre
MFR_Acre
SFR_Acre
VAC_Acre
0.24
1.69
0.72
0.72
*
*
*
1.66
3.88
3.76
- 0.1a
*
*
*
*
Model Adjusted R-Squared
a = Zoned, not buliding area
* = Sigificance < 0.05
0.84
*
0.93
*
Water Use: Structures & Demographics
200
160
120
80
40
0
1,000
200
Water Consumption Per Household
Water Consumption Per Household
Water Consumption Per Household
200
160
120
80
40
0
2,000
3,000
0
4,000
4
6
8
120
80
40
0
Av erage Household Income
150,000
4
8
12
16
Number of Houses Per Acre
200
160
120
80
40
0
100,000
40
0
Water Consumption Per Household
160
50,000
80
0
10
200
Water Consumption Per Household
200
0
120
Av erage Number of Rooms
Av erage Building Area (SQFT)
Water Consumption Per Household
2
160
160
120
80
40
0
0
20
40
60
80
100
Percentage of Households with College Education
0
20
40
60
80
100
Percentage of Houses Occupied By Owner
Water Use: Residential Structures
200
160
120
80
40
160
120
80
40
0
2,000
3,000
0
4,000
200
Water Consumption Per Household
200
Water Consumption Per Household
4
6
8
160
120
80
40
0
10
0
Av erage Number of Rooms
Av erage Building Area (SQFT)
160
2
Building Variable
160
4
8
12
16
Number of Houses Per Acre
Standard
TCoefficient
Error
Statistic
200
Water Consumption Per Household
0
1,000
200
Water Consumption Per Household
Water Consumption Per Household
Water Consumption Per Household
200
160
120
120
Constant
101.834
7.869
12.901
80
80
Building size (per 1000) 80
0.783
0.033
23.361
Density (hh/acre)
-4.904
1.0140
-4.858
40
40
Years built (per year)
-0.171
0.109
-1.56
0
0
0
20
50,000
100,000
150,000
20
40
60
80
100
0
20
40
60
80
100
N0 = 398
block
groups,
R =
0.71
Av erage Household Income
Percentage of Households with College Education
Percentage of Houses Occupied By Owner
All Significant at P < 0.001
120
Water use per household
High water use neighborhood
Water use (acre/ft)
Residential water
use is affected by
land development
characteristics
Building size (ft2/1000)
Low water use neighborhood
Parcel Level SFR Water Use
Low
High
Low
High
Results: Temperature Analysis 2
• Seasonal patterns are strong
– Threshold increase in water use occurring at 690F
Seasonal Water Use vs. Hi Temp
90.00
HH Average Water Use (CCF)
80.00
70.00
y = 0.6778x - 15.149
R2 = 0.4456
60.00
50.00
y = 0.0058x + 23.72
2
40.00
R = 3E-05
30.00
20.00
10.00
0.00
30.00
40.00
50.00
60.00
70.00
80.00
Daily High Temp (F)
Summer
Linear (Summer )
Winter
Linear (Winter)
90.00
100.00
110.00
Results: Temperature Analysis 1
• Temperature as predictors of water use
– Noisy data linking temperature to water use
• Water use becomes more varied as daily max temperatures rise
– Maximum and weekend explain almost 75% of water use
Select Household Water Use and High Temp
Select Households Water Use Profiles
(July - Dec, 2001)
80
Beta
Coefficient
28.182
0.524
3.882
Weather
Parameter
2.00
Water Use (CCF)
70
60
Water Use (CCF)
1.50
Constant
1.00
Maximum
(F)
0.50
Weekend
(0/1)
50
y = 0.2907x + 8.2004
40
y = 0.1102x + 16.388
30
y = 0.3693x - 14.056
20
y = 0.2505x - 6.2052
10
0.000
50
40
2
Adjusted
60
70 R 80
60
50
5180202063 - 600 sqft
Linear (5180202063 - 600 sqft)
70
80
90
Daily high temp (F)
High Daily Temp (F)
90
0.74110
100
100
3350172074 - 1146 sqft
4760360054 - 1544 sq ft
1470466044 - 2607 sqft
Linear (1470466044 - 2607 sqft)
Linear (4760360054 - 1544 sq ft)
Linear (3350172074 - 1146 sqft)
110
HH1
HH2
HH3
*
HH5
*
HH6
HH4
Understanding Water Use Dynamics
Driver
Temperature
Maximum, Timing
Structural
Pattern
Building area, density, age
Effect
Total Water
Consumption
Household, Neighborhood
Climate
Scenarios
Alternative
Development
Patterns
Predications of
Climate-Induced
Water Demand
Seasonal Changes in Water Use
• Change in water use between coldest day and
warmest day (2001, census block geography)
Future Climate Scenarios
• Climate Impact Group’s IPSL_CM4 A2 used to
examine model accuracy for 2001: suggest 98%
accuracy
Future Climate Scenarios
• What is the potential change in SFR water use
based on down-scaled climate models of the PNW?
– Modeling water use (based on predicted
temperature) through 2020
Average Houshold Water Use and Date Projections
(July - Dec, 2020)
60
40
30
20
10
0
7/1
7/8
7/15
7/22
7/29
8/5
8/12
8/19
8/26
9/2
9/9
9/16
9/23
9/30
10/7
10/14
10/21
10/28
11/4
11/11
11/18
11/25
12/2
12/9
12/16
12/23
12/30
Water Use
50
Date (dd/mm)
Forecasting Changes in Water Use
• Using only results from temperature data
– Percent change in water use from 2001 – 2040 (IPSL)
Implications of Climate Change on
Residential Water Use
•
•
Scenario: An 80F day in a development containing 100 new houses
(4K sqft)
–
For an average house (~ 2K square feet), each additional 1000
square feet of SFR house creates an additional water demand of
82 gallons
–
Compounded by temperature
• These houses are about 2000 square feet larger than the
average house in the Region, and 10 degrees above the 700F
degree threshold.
• The combined impact of this new development is
approximately 20,900 gallons per day, which is approximately
the amount of water stored in a typical household swimming
pool
The role of Land Use Planning
–
An additional 25% reduction in the average building size – a
reduction from 2,800 ft2 to 2,100 ft2 in the study region –
• Annual saving of 6.6 million gallons of water.
• For a new residential development of 100 SFR units, annual
decrease of 651 million gallons per year -- 1000 Olympic size
swimming pools.
Implications of Coupling Water and Land Use
Management
• Short term: Urban form, such as
housing density and building size,
need to be considered when
planning for water demand.
• Long term: We need to develop
mechanisms for better
coordinating land and water use
management, especially in the
urban and urbanizing areas
• Linking climate-induced water
demand with expected changes
to water supplies will improve
society’s capacity for predicting
future water needs
Research
• Develop models that link
alternative urban land use patterns
and temperature scenarios to
predict differences in water demand
• Identify thresholds of population
growth, land use patterns and
temperature regimes that combine
to
• Compare and contract findings
from Portland (OR) with other urban
regions in the United States
• Collaborate with regional managers
to identify and evaluate policies for
better coordinating water and land
use policies
Papers for Reference
• Shandas, V., and GH Parandvash (2009). Integrating Urban Form and
Demographics in Water-Demand Management: An empirical case study
of Portland, OR. Environment and Planning B, Vol 35 (1)
• Chang, H., GH Parandvash, and V Shandas (2010). Spatial Variations of
Single Family Residential Water Use in Portland, Oregon, Urban
Geography
• Shandas, V., M Rao, and MM McGrath (under review). The implications
of climate change on residential water use: A micro-scale analysis of
Portland (OR). Journal of Water and Climate Change
Acknowledgements and Contact Info
Portland State University
Heejun Chang
Geography
Meenakshi Rao
Environmental Science and Management
Portland Water Bureau
G. Hossein Parandvash
Principal Economist
Contact Information
Vivek Shandas
vshandas@pdx.edu
http://web.pdx.edu/~vshandas/