US Army Corps of Engineers BUILDING STRONG ® Omaha inflow forecasts based on snow regressions Second Annual Missouri Basin River Forecaster’s Meeting 15-17 November 2010 Carrie Vuyovich and Steven Daly ERDC-CRREL Cold Regions Res. and Engr. Lab. (CRREL) U.S. Army Corps of Engineers November 2010
Omaha inflow forecasts based on snow regressions Second Annual Missouri Basin River Forecaster’s Meeting 15-17 November 2010. Carrie Vuyovich and Steven Daly ERDC-CRREL Cold Regions Res. and Engr. Lab. (CRREL) U.S. Army Corps of Engineers November 2010. Project overview. - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
US Army Corps of EngineersBUILDING STRONG®
Omaha inflow forecasts based on snow regressionsSecond Annual Missouri Basin River Forecaster’s Meeting15-17 November 2010
Carrie Vuyovich and Steven DalyERDC-CRRELCold Regions Res. and Engr. Lab. (CRREL)U.S. Army Corps of Engineers
November 2010
BUILDING STRONG®
Project overview
Improve spring forecasts for reservoir inflow at 6 major basins
Current method: Monthly forecasts for 1 July inflow volume
► 1JAN, 1FEB, 1MAR, 1APR Developed in late 1980s Multiple linear regression, based on:
► Basin average SWE► Average spring (Apr – Jun) precipitation (5.73)► Antecedent flow (Oct – Nov)► Holdouts
BUILDING STRONG®
Project overview
To improve forecasts: Improve estimate of SWE Update regression
equations with recent data Look for climate change
trends in earlier snowmelt, rising temperatures, changes in precipitation
BUILDING STRONG®
Canyon Ferry
BUILDING STRONG®
Additional SWE regression parameters• Elevation• Latitude• Aspect• Regional Slope
Improve basin SWE estimate
Methods to calculate basin SWE volume• Station Average• IDW• IDW adjusted for elevation trend in SWE
BUILDING STRONG®Average Max annual SWE
0
n
ii
BasinArea SWEn
n = number of reporting stationsSWEi = SWE depth reported at SNOTEL station
Station Average
Total Basin SWE =
BUILDING STRONG®
20 ,
20 ,
1
ni
i i jj n
i i j
SWEd
SWE
d
SWEj = Average SWE depth in 1 km grid celldi,j = distance between station, i and grid cell, j
Interpolation using Inverse Distance Weighting
Total Basin SWE =
k = number of grid cells in basin
0
k
cell jj
Area SWE
BUILDING STRONG®
20 ,
20 ,
( )
1
ni SWE i
i i jj SWE jn
i i j
SWE L Elevd
SWE L Elev
d
LSWE = estimated rate of change of SWE with elevationElevi = Elevation of SNOTEL stationElevj = Elevation of1 km grid cell
Interpolation using IDW, adjusting for elevation
Total Basin SWE =0
k
cell jj
Area SWE
k = number of grid cells in basin
BUILDING STRONG®
BUILDING STRONG®
Average 1 Apr SWE by Elevation
April 1 - Average
ALBRO LAKE
BARKER LAKES
BASIN CREEKBEAGLE SPRINGS
BEAVER CREEK
BLACK BEAR
BLOODY DICK
BOULDER MOUNTAINBRACKETT CREEK
CALVERT CREEK
CARROT BASIN
CLOVER MEADOW
DARKHORSE LAKE
DIVIDE
FROHNER MEADOW
LAKEVIEW RIDGE LEMHI RIDGE
LONE MOUNTAINLOWER TWIN
MADISON PLATEAU
MULE CREEK
PETERSON MEADOWSPICKFOOT CREEK
PORCUPINE
ROCKER PEAK
S FORK SHIELDS
SACAJAWEA
SADDLE MTN
SHORT CREEK
SHOWER FALLS
TEPEE CREEK
TIZER BASIN
WARM SPRINGS
WEST YELLOWSTONE
WHISKEY CREEK
y = 0.4565x - 669.09R2 = 0.2955
0
200
400
600
800
1000
1200
1900 2000 2100 2200 2300 2400 2500 2600 2700 2800
Elev (m)
SWE
(cub
ic m
eter
s)
Other attributes showed little or no correlation with SWE, including:• Latitude• Aspect • Regional Slope
Calculated total basin SWE using SSM/I passive microwave data. SWE was significantly underestimated, most likely due to the saturation level of the satellite instrument.
Variables:• 1 Apr SWE• Antecedent Flow (Oct-Nov)• Previous years Max SWE• Apr – Jun Total Precipitation• Average Apr-Jun Total Precipitation (5.79 in)• Previous years total annual precipitation• Thiessen-weighted precipitation
BUILDING STRONG®
Average Annual Inflow volume 63085409R2 Std Error
Station Average SWE 1 Apr SWE 0.62751Apr SWE + Antec Flow 0.6869 0.247
1 Apr SWE + Last year Max SWE + Spring Precip 0.8228 0.1891 Apr SWE + Antec Flow + Spring Precip 0.8853 0.152
1 Apr SWE + Antec Flow + Prev yrs total Precip 0.7149 0.240
IDW SWE 1 Apr SWE 0.64301Apr SWE + Antec Flow 0.7028 0.241
1 Apr SWE + Last year Max SWE + Spring Precip 0.8322 0.1841 Apr SWE + Antec Flow + Spring Precip 0.8970 0.144
1 Apr SWE + Antec Flow + Prev yrs total Precip 0.7311 0.233
IDW-Elev SWE 1 Apr SWE 0.51511Apr SWE + Antec Flow 0.5890 0.283
1 Apr SWE + Last year Max SWE + Spring Precip 0.7468 0.2262 Apr SWE + Antec Flow + Spring Precip 0.824 0.188
1 Apr SWE + Antec Flow + Prev yrs total Precip 0.6353 0.271
Estimation of SWE Further analysis using SCA Station weightingMultiple linear regression Climate indices/NWS long-term forecast for additional information on spring
precipitation Holdouts Calculate significance of parametersClimate Trends Timing of precipitation Regularly incorporate additional data to regression analysisOther basins Similar SWE analysis Update regressionsNew forecast techniques ready to test by 1 Jan 2011.