a short primer on the prism2 dss
TRANSCRIPT
A Short Primer on the PRISM2 DSS
Paul Conrads, USGS
Drought and Salinity Intrusion Workshop December 14, 2011
Memory Lane
2002
Quiet Desperation Bull Creek Intake
Fast Forward
FERC
Re-licensing
PRISM: Controlled Releases Effect on Salinity Intrusion
What causes the large salinity intrusions along the coast?
What is the minimum flow to control salinity intrusion?
Is there enough water in NC to control salinity intrusions?
Data Mining
The physics is manifested in the data
Learn/quantify important cause-effect relations
Data driven models
are “virtual processes”
evaluate alternatives
Description of the Salinity Models
Cascading models
1st Simulates daily SC response
2nd simulates hourly SC response
Decomposed
Water-level signal
Input Data for Models
Riverine Flows
Pee Dee*
Little Pee Dee
Lynches
Black
Waccamaw
*User controlled input
Tidal Forcing
Mean Water level
Tidal Range
Little River Inlet
Note: Specific
conductance is not
used as an input
Freshwater
Saltwater
SC Model – Hagley Landing
Black – measured Gray – simulated
Model Performance Pawleys Island
Decision Support Systems
Models embedded in an Excel application
Integrates:
Historical database
Models and model controls
Streaming graphics
Optimization routines
Simulation outputs
Excel levels the technical playing field
DSS Architecture
Graphical User Interface
Application Controls Simulation controls –
period and time step
Flow control –
percent historical or
constant
Input data –
Wind speed
and direction
Water levels
Flows
Graphical Display
Input data - wind,
water level, and flow
Specific Conductance
– daily predictions
Specific Conductance
– hourly predictions
Actual data –
black
Model
prediction –
red
User
specified
condition –
green
Difference
b/w user and
actual - gray
Looking Inside the Black Box
3D response surfaces Surface created by
ANN model
“Unseen” variables set to constant value
Manifestation of historical behavior of system
Insight to the process dynamics or physics
What Conditions Cause Large Intrusions?
High
water
Low
water
3-D Response surface – Q, XWL, and SC
Low Coastal Water Level High Coastal Water Level
Convergence of Conditions
PRISM2: Threaten Intakes along SE Coast Due to Climate Change
• Sea-level rise
• Changing
streamflow
patterns
• Inputs from
GCMs
PRISM > PRISM2
User control of uncontrolled and controlled rivers
User control to bias sea levels
User specified hydrographs for rivers
PRISM2: User Controls
Riverine Flows
Pee Dee*
Little Pee Dee*
Lynches*
Black*
Waccamaw*
Tidal Forcing
Mean Water
level*
Tidal Range
Little River Inlet
* User controlled inputs
Freshwater
Saltwater
Conceptual Model
Precipitation
Temperature
Global & regional
circulation models
Gridded rainfall
input to watershed
model Watershed model
<Sea-level rise
Salinity intrusion
model
Data Stream
Precipitation
Temperature
Geo Data Portal
GCM downscale
data
HSPF
Watershed model
<Sea-level rise
PRISM2
• Upload shape file of study area
• Climate data for each sub-
watersheds
24
• Configure retrieval statistics/format
• Pull data for any number of
models and scenarios
• User-defined climate projection
period
25
PRISM2: Splash Page
Run Page
Input Set Points Page
User Defined Hydrographs
Output
• 132 output columns
• All input values
• All input setting
• Measured values at each
gage
• Model Output
• Measured
• User specified
• Delta
• Measured + delta
• Daily & hourly values
Results – Sea-level Rise Grand Strand – Pawleys Island Gage
Simulated a 0.5, 1.5, 1.5, 2.0, 2.5 and 3.0 ft SLR
14 year simulation
Comment on extrapolation
Large range in historical data
annual coastal water levels
Lower ranges of SLR (<2.0 ft) within range of historical data
Output Example 25% Reduced Flows
Flows and Delta SC
Sea-level rise projections Springmaid Pier
Local Trend
Modified NRC-III
Modified NRC-II
Modified NRC-I
Analysis based on USACE methodology
Multiple Model Runs Results - SLR Grand Strand – Pawleys Island Gage
14 year simulation
0.5 ft incremental SLR up to 3 ft
Results – SLR & Reduced fow Grand Strand – Pawleys Island Gage
Reduced flows by 5% up to %25
Figure 1. Study area including the Pee Dee and Waccamaw River Basins and Atlantic Intracoastal Waterway in South Carolina.
Pee Dee Flows Waccamaw Flows
Tidal range 3 ft
Tidal range 4 ft
Freshwater slug With sea-level rise, appears
that it will push more
freshwater to the north.
Water level rises in the
“slug” and forces more
water to the north and
pushes out the salt.
Pee Dee,
Wacammaw and
AIW
System Behavior
South End Response
North End Response