2005 Year End Floods

Southwest Oregon

December 22 - 31, 2005




Heavy rains between Christmas and New Year's caused flooding from Western Oregon to Northern California.  This report covers the Southwest Oregon flood events for this period - stretching from the Southern Oregon Cascades to the Southern Coastal Oregon range and to the Southwest Oregon Coast.  Storm precipitation totals topped 20 inches at several locations in Southwest Oregon between December 22 and January 1 (Tables 1).  Rivers in the Umpqua, Rogue, and Coquille River Basins overtopped their banks and inundated adjacent land.  The most severe flooding occurred in the Umpqua River basin where 500 people were evacuated in the vicinity of Deer Creek, and in the Rogue River basin at Grant's Pass where river heights approached major flood levels.  The Rogue River at Grant's Pass crested at 24.09 feet at 0728Z December 31st, one-half foot below major flood classification.  Douglas, Curry, Coos, Jackson, and Josephine Counties were affected by flooding.  Jackson County declared a state of emergency to aid in flood mitigation.  Damage, as reported by local papers, included numerous road closures due to excess water and landslides, and property flooding.


Meteorological Summary

Typical seasonable rains (and mountain snow) incrementally primed westside upper soil zones across western Washington and western Oregon through early to mid December.  Southwestern Oregon coast range and Washington Olympic Mountain headwaters received local 9 to 10 inch maximum accumulations through mid December (Figure 1).   Prior to December 19th, Southwest Oregon experienced near normal precipitation for the month of December and the 2006 Water Year (October - December).  In contrast the majority of the Columbia basin experienced essentially below normal for December up to the 19th, and slightly below normal on average for the water-year (from October 1, 2005):


   Columbia Basin Seasonal Precipitation Division Averages

  Northwest River Forecast Center


 (Weekly Summary: 12Z Dec 1 - 19, 2005)

   DIVISION                ..DEC TO DAY 19..    ...OCT - DEC...

                           OBSD   DEP  PCT AV   OBSD   DEP  PCT AV


   COLUMBIA ABOVE COULEE   0.70  -1.15  38.     6.13  -0.45  93.

   SNAKE RV AB ICE HARBOR  0.98  -0.22  81.     4.46   0.17 104.

   COLUMBIA AB THE DALLES  0.84  -0.98  46.     5.47  -0.84  87.

   ROGUE/UMPQUA            3.64   0.14 104.    13.74   1.89 116.



Figure 1. Six Day Accumulative Precipitation - December 16-21, 2005


Near seasonable early to middle December weather was followed by a persistent very wet pattern resulting from an extended long wave length upper level trough and relatively uninterrupted jet-stream that advected a well defined continuous plume of 1 to 1.5 inch plus precipitable water from the vicinity of the Philippines (near the base of the long wave trough) northeastward on a track into the Pacific Northwest (Figure 2).

Figure 2. Global Integrated Cloud Liquid Water - December 26-28, 2005



This fairly steady state late December cross-Pacific moist atmospheric pattern resulted in frequent training of unsettled wet weather, particularly into NWRFC coast range and Cascade headwaters.  These storms had their greatest effect on NWRFC westside basins, however, this scenario gradually extended scattered 1 plus inch daily rains east of the Cascades as well (mainly into middle Snake River drainages across south-central Idaho).

Throughout this late December event period storms were frequently accompanied by both strong low level (50 knot plus) and upper level (130 knot plus) jet streaks as individual baroclinic waves initially pushed onshore into the Pacific Northwest.


From December 22 to December 27, an almost uninterrupted series of Pacific storm systems slowly sagged a pattern of intensifying rains southward along the coast - starting across northwestern Washington (Figure 3, forecast day-1 12/22/05 QPF) and then gradually stalling across southwestern Oregon (Figure 4, forecast day-1 12/27/05 QPF).  Local moderate rains then spread across much of western Oregon and western Washington December 28th to the 29th, but heavy rains soon redeveloped across south-western Oregon (and to some extent across the NWS Portland forecast area) on December 30-31 (Figure 5, forecast day-1 12/30/05 QPF).  Slowly diminishing onshore surges of continuing precipitation at the end of this event gradually shifted back northward, aiming occasional local moderate to heavy rains mainly into northwestern Washington on January 1, 2006 - thus curtailing precipitation driven surface runoff input to flooding main stem flows in the NWS Medford forecast area.


Figure 3.  Forecast day-1 12/22/05 Quantitative Precipitation Forecast



Figure 4.  Forecast day-1 12/27/05 Quantitative Precipitation Forecast

Figure 5.  Forecast day-1 12/30/05 Quantitative Precipitation Forecast


The prolonged late December southwesterly fetch of slowly modifying Western Pacific moist tropical air resulted in not only frequent impulses of locally heavy westside rains, but also the advection of high dew point warm air across much of the NWRFC forecast area (Figure 6).  As a result, freezing levels occasionally climbed up to 8000-11000 across southern Oregon and southern Idaho basins.  This lofting of freezing levels often coincided with surges of the heaviest observed rains, thus also contributing to locally significant translation of pre-existing low to mid elevation snow packs into quickly available surface runoff.  This was the case for Middle Snake River tributaries in the southwestern Sawtooth Mountains, primarily north-northwest of Boise.  These basins experienced sharp local hydrologic runoff responses as a result of this combination of warming with rain on pre-existing snow.

Figure 6. Visible Satellite Image - December 27, 2005


By end of December 2005 observed precipitation across the Columbia basin climbed above normal for December and the water year (October - December 2005).  Precipitation in the Rogue/Umpqua area reached  212 percent of normal for December.  Monthly precipitation for December 2005 for the Pacific Northwest is shown in Figure 7.




Columbia Basin Seasonal Precipitation Division Averages

  Northwest River Forecast Center

 (Monthly Summary: - Dec 2005)

   DIVISION               ..DEC TO DAY 31..    ....OCT - DEC....

                           OBSD  DEP PCT AV    OBSD   DEP  PCT AV


   COLUMBIA ABOVE COULEE   2.50  -0.52  83.     7.93  0.18  102.

   SNAKE RV AB ICE HARBOR  3.60   1.63 183.     7.11  2.06  141.

   COLUMBIA AB THE DALLES  3.42   0.44 115.     8.07  0.60  108.

   ROGUE/UMPQUA           12.13   6.41 212.    22.23  8.17  158.





Figure 7. Monthly Precipitation Percent of Normal for December 2005




The heaviest precipitation fell across the NWRFC area between December 22, 2005 and January 1, 2006.  Several locations in the Southwest Oregon received 15 to 26 inch rain accumulations during this period (Figure 8).  For climatic comparison this is nearly twice the amount for the normal precipitation totals for some places in Southwest Oregon.  The following is monthly normal data from the Western Regional Climate Center: BROOKINGS 2SE 13.12 inches; ILLAHE   14.60 inches; GOLD BEACH RANGER STATION 13.78 inches.  Heaviest precipitation days occurred on December 27th and 30th  (Figures 9 and 10, Table 4).





Figure 8.  Storm Total Precipitation December 22 - January 1st



Figure 9.   24 Hour Observed Precipitation - December 27, 2005



Figure 10.   24 Hour Observed Precipitation  - December 30, 2005





                                          Storm Precipitation


                                                                                    Daily          Daily      Storm Total

NWS ID            NAME                   QPS SITE          Dec 27       Dec 30    Dec 22-Jan 01


LAFO3       Lawson RAWS                                         5.63     6.03    33.58

VLZO3       Valsetz                                                      3.83     2.63    28.25

BKFO3       Bald Knob 2                                             4.46     4.65    28.03

ELKC1        Elk City                                                    4.51     4.84    26.84

LAMO3       Laurel Mountain                                       3.56     2.19    26.60

CVFO3        Calvert Peak RAWS                                5.19     2.29    25.00

ILHO3         Illahe                                QPS                 4.00     4.80    23.70

GASC1        Gasquet RS                                             2.79     4.00    23.28

CVJO3        Cave Junction                                          3.47     4.13    23.08

GLBO3       Gold Beach RS                                        3.64     3.92    22.37

GOLO3       Gold Beach                                              3.63     3.91    22.24

4BK            Brookings                         QPS                 3.20     3.86    21.70 

KRBO3*      Kirby                                                         3.21     3.79    21.57

SITO3         Sitkum 1E                                                 3.57     3.73    20.62

LMDO3       Little Meadows SNOTEL                          1.89     4.40    18.57

POWO3      SF Coquille R at Powers                          2.91     2.96    17.68

MTHO3       Mt Hood Test SNOTEL                            2.48     4.13    16.83

MRFO3       Marion Forks SNOTEL                             2.00     4.19    14.96

GSVO3       Galesville Reservoir                                  2.52     1.57    14.00

SXT            Sexton Mountain                                       2.40     2.50    13.91

CRLO3       Crater Lake                                               1.89     2.59    13.60

SCFO3       Salt Creek Falls SNOTEL                         1.70     4.40    12.99

WLMO3*    Williams                               QPS              1.63     2.49    12.46 

OTH           North Bend                          QPS               1.75     1.70    12.41     

SRSO3      Steamboat Ranger Station  QPS               1.89     3.79    12.34  

RDLO3      Cow Creek nr Riddle            QPS              2.50     2.00    11.90 

TILO3         S Umpqua River at Tiller                           1.82     2.36    10.69

BLBO3       Black Butte 1 N                                         1.20     2.36     9.48

BUFO3       SF Big Butte Cr nr Buttefalls                       .89     2.50     8.46

EMIO3        Emigrant Cr nr Ashland                              .57     3.14     7.29

MFR           Medford                               QPS                .93     2.02     5.95   

* Estimated

Table 1. Southwest Oregon Storm Total Precipitation (from NWRFC Daily Precipitation Quality Control)



Flood Summary

Frequent, moisture rich, Winter storms on their own are not unusual in the Pacific Northwest.  What makes this instance unusual is the prolonged and consistent nature of the upper level jet stream that developed stretching from the Philippines to the Pacific Northwest.  This atmospheric pattern brought frequent December storms, with high wind jet cores and high precipitable water content, to the Northwest region.  In early and mid December, these Pacific systems primarily eclipsed Northwest Washington and the Olympic Peninsula.  By December 22nd, the jet pattern sagged south; the plume of high precipitable water was now focused on Southwest Oregon and Northern California.


Prior to December 19th, Southwest Oregon experienced near normal precipitation for the month of December and the 2006 Water Year (October - December); 104 and 116 Percent of Normal precipitation.  By end of month, precipitation amounts for the Rogue and Umpqua area rose to 212 and 158 Percent of Normal for the month and 2006 Water Season. 


The heaviest 24 hour precipitation periods occurred on December 27th and December 30th.   This is coincidental with river rises on December 28th and December 31st.


Freezing levels remained high during the event therefore rain fell at most elevations.  NRCS Snotel sites within the Rogue and Umpqua basins reveal that snow water equivalent measurements at sites below 5,200 feet fluctuated only a few tenths of an inch from December 22 - 31st.  Snotel sites above 6,000 feet showed little change from December 22 - 30th, then increase approximately 1.5 inches of snow water equivalent from the 30th to 31st.  This signifies that snow melt or rain on snow instances did not contribute greatly to the magnitude of the flood event (Table 2). 








 Data Site Name       (Ft)                      %                      %

                             Current  Average  Avg  Current  Average  Avg


Dec 22

 BIG RED MOUNTAIN     6250     8.2      9.9     83     22.7    18.2   125

 BIGELOW CAMP         5120      .9      5.0     18     28.7    23.1   124

 DIAMOND LAKE         5315     4.4      6.4     69     18.7    17.2   109

 FISH LK.             4665     3.6      5.3     68     19.1    14.9   128

 KING MOUNTAIN        4000      .1      2.1      5     22.1    21.6   102

 SEVENMILE MARSH      6200    13.5     11.2    121     26.3    22.1   119

                                              -----                  -----

           Basin wide percent of average        77                    118

Dec 28

 BIG RED MOUNTAIN     6250     8.9     10.9     82     26.2    19.9   132

 BIGELOW CAMP         5120     1.3      5.5     24     35.2    25.2   140

 DIAMOND LAKE         5315     3.7      6.9     54     22.1    18.8   118

 FISH LK.             4665     3.0      5.8     52     20.6    16.2   127

 KING MOUNTAIN        4000      .0      2.4      0     27.0    23.6   114

 SEVENMILE MARSH      6200    14.9     12.5    119     29.6    24.1   123

                                              -----                  -----

           Basin wide percent of average        72                    126

Dec 31

 BIG RED MOUNTAIN     6250     11.9   11.4    104     31.9    20.7   154

 BIGELOW CAMP         5120      1.0    5.7     18     42.4    26.2   162

 DIAMOND LAKE         5315      5.3    7.2     74     26.8    19.5   137

 FISH LK.             4665      4.1    6.1     67     23.5    16.9   139

 KING MOUNTAIN        4000       .0    2.5      0     33.2    24.6   135

 SEVENMILE MARSH      6200     17.7   13.2    134     35.9    25.2   142

                                            -----                  -----

          Basin wide percent of average        87                    146


Table 2. NRCS Snotel observed snow water equivalent (Rogue and Umpqua Basins)





The flood portion of this report focuses on rivers where the National Weather Service provides river stage forecasts.  Figure 11 is a display of locations where the National Weather Service produces river forecasts.


Flooding occurred in all of the major river basins of Southwest Oregon; along the Umpqua, Rogue, and Coquille river basins.  Hydrographs are provided to show the relative magnitude of the event at all locations where flood level was reached.  Plots are made from preliminary data that is subject to change.


Figure 11.   NWRFC Southwest Oregon River Forecast Site Locations and  NWS Quantitative Precipitation Sites.



Rogue River

Lost Lake Reservoir, 30 miles northeast of Medford in the headwaters of the Rogue River, is owned and operated by the Corps of Engineers (COE).  The COE references NWRFC forecasts for management of the project.  Final COE discharges are incorporated into NWRFC forecasts prior to forecast issuance.


Inflow into Lost Lake Reservoir exhibited a double peak in late December.  Peak inflows for December 29th and 31st were approximately 8,000 cfs and 15,000 cfs (Figure 12).


Figure 12.  Inflow Hydrograph for Lost Lake Reservoir.


Figure 13. Discharge Hydrograph for Lost Lake Reservoir.



The Corps of Engineers cut back Lost Creek reservoir releases to minimum outflow at the time the Rogue River near Eagle Point (forecast area downstream) was cresting and thus lowered the actual Eagle Point crest and subsequent Rogue River forecast area sites by several feet (Figure 13 and 14.)


Figure 14.  Flood Hydrograph for the Rogue River near Eagle Point.


With Lost Lake Discharges at minimum flow, most of the water driving the Rogue River near Eagle Point to flood stage came from Big Butte and Eagle Creeks (Figure 14.)


 Figure 15.  Flood Hydrograph for Bear Creek at Medford.



Runoff from the cities of Ashland and Medford combined with the runoff from the surrounding hills to produce a peak almost two feet above flood stage at Medford (Figure 15).


Figure 16.  Flood Hydrograph for the Rogue River at Raygold.



Flood waters from Bear Creek and Eagle Point joined to produce a flood four feet over flood stage at Raygold (Figure 16).   


Figure 17.  Flood Hydrograph for the Rogue River at Grants Pass.



The flood peak at Grants Pass peaked a few hours after Raygold approximately four and half feet over flood stage.  Grants Pass and Raygold benefited by the operation of Lost Creek Reservoir.  Without the reduction in discharges at Lost Creek Reservoir it is likely that Grants Pass would have exceeded major flood levels (Figure 17).




Figure 18.  Flood Hydrograph for the Rogue River near Agness.



Rain prior to December 27th brought river stages at the lower end of the Rogue River near Agness up to near 10 feet on the 22nd.  Agness stages remained high until the heavy rains of December 27th and 28th pushed river levels upward to a crest of approximately 24 feet.   River levels no more fell below flood stage on the 29th before December 30th and 31st rains forced a second peak of about 31 feet; 14 feet over flood stage on the Rogue River near Agness (Figure 18).


Umpqua River

Figure 19.  Flood Hydrograph for the South Umpqua River at Tiller


Figure 20.  Flood Hydrograph for the South Umpqua River near Riddle



Figure 21.  Flood Hydrograph for the South Umpqua River near Brockway.



Figure 22.  Flood Hydrograph for the South Umpqua River at Roseburg


Figure 23.  Flood Hydrograph for Deer Creek near Roseburg

Figure 24.  Flood Hydrograph for the Elk River near Drain


Figure 25.  Flood Hydrograph for the Coquille River at Myrtle Point

Figure 26.  Flood Hydrograph for the North Fork Coquille River at Myrtle Point


Figure 27.  Flood Hydrograph for the Coquille River at Coquille




Forecast Verification


Precipitation Forecasts


The Hydrologic Prediction Center, NOAA (HPC) produces six hourly quantitative precipitation forecasts (QPS) for use in the NWRFC river forecast model.   Figure 11 shows the location of the eight QPS points in southwest Oregon.  HPC precipitation forecasts are reviewed and possibly adjusted at the Medford WFO and the NWRFC prior to being incorporated into the hydrologic model.  Figures 28  through 35 show observed precipitation and the forecasted precipitation one, two, and three days prior to observed.   Note the Williams precipitation gage was not operational during the event; observed precipitation at Williams is produced from the NWRFC daily precipitation quality control process. 


Figure 28. Forecasted and Observed 6 Hour Cumulative Precipitation for North Bend, Oregon.


Figure 29. Forecasted and Observed 6 Hour Cumulative Precipitation for Illahe, Oregon.


Figure 30. Forecasted and Observed 6 Hour Cumulative Precipitation for Brooking, Oregon.

Figure 31. Forecasted and Observed 6 Hour Cumulative Precipitation for Steamboat Springs, Oregon.


Figure 32. Forecasted and Observed 6 Hour Cumulative Precipitation for Riddle, Oregon.


Figure 33. Forecasted and Observed 6 Hour Cumulative Precipitation for Sexton Summit, Oregon.

Figure 343. Forecasted and Observed 6 Hour Cumulative Precipitation for Williams, Oregon.


Figure 35. Forecasted and Observed 6 Hour Cumulative Precipitation for Crater Lake, Oregon



In general the QPS graphics reveal the portending of three waves of precipitation.   The first moderate event occurring in and around the 25th, a second heavy precipitation event confined somewhat to coastal and more mountainous Southwest Oregon areas on the 27/28, and a third yet more substantial precipitation event Southwest Oregon basin wide on the 30th.   In general,  the 25th event did not occur or marginally happened on the 26th.  The second and third events were well forecasted in that moderate to heavy precipitation occurred on the 27th, 28th , and again on the 30th with peak precipitation timing within six hours of observed timing in all three major river basins; Rogue, Umpqua, and Coquille.


A closer look at the December 27/28 precipitation event shows a bias toward over forecasting the more coastal QPS sites (North Bend, Illahe, and Brookings) and a good grasp (within a couple tenths of an inch) of the precipitation event two days prior for the inland and more mountainous QPS sites (Steamboat Springs, Riddle, Sexton Mountain, Crater Lake). 


The December 30th event in general was within a half-inch of observed two days in advance for many of the Coastal QPS sites and Crater Lake with the exception to Illahe where 1.58 was predicted and 2.70 was observed.  Inland, however,  in particular in the Umpqua and Rogue areas,  the event was under forecasted.  The following is a table of December 30th peak observed six hourly precipitation to peak forecast six hourly precipitation within two synoptic periods.


                                     December 30th Peak Precipitation (Obs vs. Fcst)




Observed Peak 6hr Precipitation

Day 1 Fcst Peak 6hr Precipitation

Day 2 Fcst

Peak 6hr


Day 3 Fcst

Peak 6hr



North Bend


















Steamboat Springs












Sexton Mountain






Crater Lake





|     |  Under forecasting



Stream Flow Forecasts


Heavy rains on the December 27th and 30th were the primary cause for river flooding in Southwest Oregon on December 28th and 31st.  To analyze the stream flow forecasts for Southwest Oregon late December,  NWRFC products are looked at for forecast lead time to actual river flood stage and  peak stage timing and height.


To study flood forecast lead time,  flooding is separated into two separate events; one flood event on the 28th (peaking Dec 29th) and a second on the 31st (peaking Jan 01).   In the first event the heaviest precipitation was confined to the Southwest Coast and lower reaches of the Rogue, Umpqua, and Coquille river basins.  On the 31st of December heavy rains were experienced in all river basin headwaters from the Coast to the upper headwaters of the Cascade draining Rogue and Umpqua (Figures 9 and 10).


NWRFC flood forecast lead time for the December 28th event is outlined in Table 3 and in Figure 36.  For forecast locations that reached flood stage on December 28th, lead times were generally greater than 36 hours with the exception of the smaller Bear Creek drainage near Medford where a flood declaration was issued 40 minutes in advance of flood. 



Table 3. Flood Lead Time for December 28th Event.


Figure 36.  Flood Lead Time for December 28th Event.



The second flood event followed less than 48 hours after the first on December 31st.   Forecast lead time is summarized in Table 4 and Figure 37.  Long leads from 38 hours to 100 hours are seen for the lower reaches of the Rogue and Coquille River basins.  Short notice, less than an hour, occurred on the North Fork Coquille at Myrtle Point, South Umpqua River near Riddle, and contributing creeks of the Rogue and Umpqua Rivers -  Eagle Creek at Dodge Bridge, Bear Creek at Medford, Deer Creek near Roseburg, Elk Creek near Drain.

Table 4. Flood Lead Time for December 31st Event.





Figure 37.  Flood Lead Time for December 31st Event.



Peak Stage Forecasts


Peak stage forecasts were extracted from NWRFC products and plotted in reference to time of observed river peak for the events in late December.   Generally, two peaks were experienced late December (Figures 12 - 26); the first one on December 29th, the second on December 31st (observed headwater peak for second event).  In general flood forecast lead time was greater for the Coastal and Lower reaches of the Rogue, Umpqua, and Coquille River basins than the headwaters of the Umpqua.   The peak stage forecast analysis follows per major river basin:


In reviewing Rogue River peak stage forecasts from NWRFC products, in most cases, flood stage was portended 2.5 days in advance of December 29th peak stage.  Minor fluctuations in river forecasts prior to December 29th are coincidental with QPS variances.


Rogue River near flood forecasts were issued immediately following the first river peak for Dec 31st (approximately 1.5 days lead time).  Forecasts remain relatively unchanged until approximately twelve hours prior to the peak on December 31st.  Forecasts were adjusted higher due to exceedence of forecast precipitation in the basin.  Of particular note, the Rogue River at Grantís Pass crested at 24.09 feet at 0728Z December 31st, one-half foot below major flood classification














Umpqua River peak stage forecasts generally showed flood stage 2.5 days in advance of December 29th peak stage.


Umpqua flood forecast lead times for the second event was less than 12 hours.  This is attributed to the precipitation event stretching into the Umpqua headwaters.  Frequent updates occurred as precipitation exceeded forecast precipitation.









Coquille River peak stage flood forecasts were issued greater than 2.5 days in advance of December 29th peak stage.   


Coquille flood forecast lead times for the second event was less than 12 hours.  This is attributed to the precipitation event stretching into the Umpqua headwaters.  Frequent updates occurred as precipitation exceeded forecast precipitation.


Coquille River flood forecasts were issued immediately following the first river peak for Dec 31st (approximately 1.5 days lead time).  Forecasts remain relatively unchanged prior to the December 31st peak.