Transcript Hydrologic Forecasting - UW Hydro | Computational Hydrology

Hydrologic Forecasting
Alan F. Hamlet
Dennis P. Lettenmaier
JISAO/CSES Climate Impacts Group
Dept. of Civil and Environmental Engineering
University of Washington
Winter Climate of the Western U.S.
DJF Temp (°C)
NDJFM Precip (mm)
Runoff Timing in the PNW is Determined Primarily by
Winter Temperature Regimes
Normalized Precip or Runoff
3.0
2.5
Snow
Dominated
2.0
Transient
Snow
1.5
Rain
Dominated
1.0
PNW
Precip
0.5
0.0
10 11 12
1
2
3
4
Month
5
6
7
8
9
Typical Effects of Urbanization on a Small Watershed
Des Moines Creek
(developed)
Source: Booth D.B., 2000, Forest Cover, Impervious-Surface Area, and the Mitigation
of Urbanization Impacts in King County, WA
http://depts.washington.edu/cwws/Research/Reports/forest.pdf
Effects of the Pacific Decadal Oscillation (PDO) and El
Niño Southern Oscillation (ENSO) on Columbia River
Summer Streamflows
Red = Warm ENSO, Blue = Cool ENSO, Green = ENSO neutral
450000
Cool PDO
Warm PDO
400000
Warm PDO
350000
300000
250000
200000
2000
1990
1980
1970
1960
1950
1940
1930
1920
1910
150000
1900
Apr-Sept Flow (cfs)
Cool PDO
Global Surface Temperatures are
Increasing Rapidly
Weather
Forecasts
Flood Control
and
Hydropower
Management
Flood
Forecasts
1hr - 1 week
Seasonal to
Interannual
Climate
Forecasts
Seasonal
Streamflow
Volumes
Water
Resources
Management
1– 24 months
Forecast Lead Time
Climate
Change
Scenarios
Long-Range
Streamflow
Forecasts
Water
Resources
Planning
10-100 years
Schematic Diagram of a Hydrologic Forecasting System
Initial Hydrologic State
•Soil Moisture
•Snowpack
Hydrologic
Forecast:
Hydrologic Model
Future Temperature and
Precipitation
Forecast
•Streamflow
•Soil Moisture
•Snowpack
•Evaporation
Simulated Water Balance for the Pacific Northwest
250
precipitation
swe
200
active soil storage
evapotranspiration
150
100
50
sep
aug
jul
jun
may
apr
mar
feb
jan
dec
nov
0
oct
Area Average Water
(depth in mm)
runoff+baseflow
Relative Roles of Future Precipitation and Initial
Hydrologic State at Different Forecast Dates
In October
future
precipitation
dominates the
inputs to the
water balance.
900
800
700
600
500
400
In April inputs to
the water
balance from
future
precipitation
and storage are
comparable.
99%
precipitation
snowmelt
soil drainage
streamflow
evaporation
300
200
46%
Water Balance (depth in mm)
1000
100
0
oct-sept
apr-sept
Simulated Long-Term Water Balance for the Pacific Northwest
Examples of Hydrologic
Forecasting Systems
Example of a Short Time Scale Flood Forecasting System
MM5 mesoscale
atmospheric model
Estimated
Hydrologic
State
DHSVM distributed
hydrologic model
Streamflow Forecast
River Stage Forecast
Example of a Seasonal Forecasting System Based on
Regression Models
NRCS SNOTEL Network
NRCS/NWRFC Water Supply Forecasts
Hydrologic
Index
Regression
Equation
Streamflow
Volume
Example of a Seasonal Forecasting System Using a
Physically-Based Hydrologic Model
Temperature and
Precipitation
Forecast
Estimated
Hydrologic
State
VIC Hydrologic Model
UW West-Wide Seasonal Hydrologic
Forecast System
http://www.hydro.washington.edu/forecast/westwide/
Hydrologic
Forecast
Background: Forecast System Schematic
local scale (1/8 degree)
weather inputs
Hydrologic
model spin up
NCDC met.
station obs.
up to 2-4
months from
current
1-2 years back
LDAS/other
real-time
met. forcings
for spin-up
gap
soil moisture
snowpack
INITIAL
STATE
streamflow, soil moisture,
snow water equivalent, runoff
Hydrologic forecast
simulation
ensemble forecasts
SNOTEL
SNOTEL
Update
Update
25th Day, Month 0
ESP traces (40)
CPC-based outlook (13)
NCEP GSM ensemble (20)
NSIPP-1 ensemble (9)
Month 6 - 12
Climate forecasts based on ENSO predictions can provide
useful information about future streamflows with lead times up
to 12 months.
Retrospective tests
in the Columbia
River basin have
shown that during
cool or warm events,
ENSO-based
streamflow forecasts
are superior to
assumptions of
“normal” conditions
about 65 % of the
time on Oct 1
Natural Streamflow (cfs)
Natural Streamflow Columbia River at The Dalles, OR
Red = Unconditional mean
Blue = Ensemble mean
Black = 2005 Observed
Conclusions
Useful hydrologic forecasts based on weather or climate
forecasts are available with lead times ranging from a few
hours (flood forecasts) to 50 years or more (climate change
scenarios).
Many operational hydrologic forecasting systems are
currently based on statistical models, however dynamic,
physically-based tools are increasingly being used in both
academic and operational forecasting systems.
Dynamic forecasting systems based on weather or climate
models directly linked to physically-based hydrologic models
have important advantages in a rapidly evolving climate
system. Short-term forecasts based on weather models have
already reached a useful state of development, but many
challenges remain at seasonal or longer time scales.