WESTERN REGION TECHNICAL ATTACHMENT
NO. 97-13
APRIL 8, 1997


QUANTITATIVE PRECIPITATION FORECAST ACTIVITIES
OF THE NWSFO RENO DURING THE
DECEMBER 1996 - JANUARY 1997 FLOODING

Steven Goldstein - NWSFO Reno, NV


Introduction

Quantitative Precipitation Forecasts (QPFs) issued by the National Weather Service (NWS) are of vital importance during excessive precipitation events. Accurate QPFs allow River Forecast Centers (RFC's) to better predict river rises and the extent of flooding expected with these rises. During the period including 20 December 1996 through 6 January 1997, an extremely wet weather pattern developed over the Eastern Pacific and western United States. The eastern Sierra Nevada Mountains and western Nevada were hit with copious amounts of precipitation. The flooding that resulted was the worst on record over portions of the Carson and Walker River Basins. The Truckee and Susan Rivers also experienced dangerous floods that resulted in millions of dollars in damages.

The Nexrad Weather Service Forecast Office (NWSFO) in Reno, NV issues daily QPFs for Woodfords and Truckee, CA (Fig. 1). These points coincide with RFC forecast points along the Truckee and Carson Rivers. Forecasters at Reno utilize guidance from the National Center for Environmental Prediction (NCEP), (Fig. 2), computer model-derived QPFs and forecasts from the Monterey, CA forecast office (Fig. 3) (Fig. 3) in assembling local QPFs. The QPFs are issued daily by the Reno office at 12 UTC. Updates are issued as warranted. These QPFs are then incorporated into RFC streamflow models to predict stages along the Truckee and Carson Rivers. This paper examines the conditions prior to the major flood event, and reviews the QPFs issued throughout the period.


Antecedent Conditions

The December 96-January 97 flooding resulted from very heavy rainfall on a large snowpack. The resultant runoff filled reservoirs to capacity and brought the area its worst flooding since 1955. A heavy snow event from 20 December through 23 December 1996 brought several feet of snow to the river basins. As much as seven feet fell at Truckee, with three feet at Markleville, CA in the Carson River Basin. The following week a deep and moist southwest flow developed over the eastern Sierra Nevada Mountains and western Nevada. This brought heavy rain to the region during the period of 30 December 1996 through 3 January 1997. Automated sensors in the Sierra Nevada Mountains (Squaw Valley Gold Coast in eastern Placer County) reported nearly 20 in of rain in 24 h from 1 January-2 January. It was during this period that the QPFs were most important as river flooding became severe.

The synoptic pattern from 23 December 1996 through 1 January 1997 began with a weak ridge of high pressure aloft over western Nevada and the Sierra Nevada Mountains of eastern California (Fig. 4). A broad long wave trough developed over the central and Eastern Pacific which lead to the eventual weakening of the ridge. Also note the presence of a strong Hudson Bay low. The jetstream winds over the central Pacific increased to 110-170 kts along the base of the trough. Moisture was provided from the tropics as the jetstream swept near Hawaii. The remains of a typhoon (FRAN) were swept along with the jetstream.

On 29 December, a short wave rode the stream of moisture from Hawaii to California and swept into the eastern Sierra Nevada Mountains and western Nevada. Heavy rain and snow spread over the region and freezing levels began to rise (Table 1). It reached 10,700 ft late in the day on New Year's Eve. Forecasters at Reno recognized the threat of major river flooding at this time. Warm, strong winds increased snowmelt and commenced runoff at lower elevations. These winds limited precipitation amounts in the lee of the Sierra Nevada Mountains at this time.

The synoptic pattern for 31 December (Fig. 5) shows that the Hudson Bay low had moved east to near 60W, allowing the deep, moist southwest flow over the Eastern Pacific to advect into the eastern Sierra Nevada Mountains and western Nevada. A strong short wave moved through the area on New Year's Day. This resulted in locally heavy rains, gale force winds and temperatures that rose well above freezing. The afternoon of 1 January, the Reno atmospheric sounding revealed a freezing level of 11,700 ft (Table 1). Moderate to heavy rains fell over a large, saturated snowpack along the higher elevations of the Sierra Nevada Mountains. The rains soon spread into the lower elevations of the western valleys of Nevada. The resultant runoff flowed in streams from the snow pack across the high Sierra Nevada Mountains into the river basins. The affects of the resultant flooding would be felt across the region for many weeks.

On 2 January, flood crests traveled quickly off the mountains of the Sierras into the river basins. That afternoon a short wave from the Gulf of Alaska (Fig. 6) dropped into the Pacific Northwest and spread colder air across the eastern Sierra Nevada Mountains and western Nevada. Snow levels the morning of the third dropped to 6500 ft ASL. Convection associated with a cold front dropped snow levels locally to 5000 ft ASL that afternoon. Rain changed to snow at the higher portions of the basins. Runoff was reduced over the next couple of days as a colder northwest flow lowered snow levels to the valley floors by 4 January.


QPFs

QPFs for Truckee and Woodfords are examined against observed precipitation in Tables 2 and 3. Observed precipitation for the four major river basins in eastern California and western Nevada is shown in Table 4. It can be seen that during the critical period from 30 December through 3 January, QPFs issued by the NWSFO Reno verified quite well. During this period the Woodfords (Carson River) QPFs issued by Reno forecasters totaled 8.65 in. The gage at Woodfords was washed out during this storm, but the gage at Markleville (Southeast of Woodfords, Fig. 1) recorded 7.25 in during these five days. Within this same time frame, the NWSFO Reno QPFs for Truckee totaled 8.15 in. Precipitation recorded at the Truckee Automated Local Evaluative in Real Time (ALERT) gage from 30 December through 3 January was 9.10 in.

Incremental QPFs issued by NWSFO Reno also performed well during this period. 0-6 h, 6-12 h and 12-24 h QPFs for Truckee and Woodfords are depicted in Tables 2 and 3. Total precipitation was initially underestimated on 1 January for Truckee. An updated QPF issued at 11:45 AM PST that morning performed substantially better (Table 2) and (Table 2A). A similar analysis of the QPFs issued for the Carson River Basin was performed by examining 6 h SNOTEL data from Spratt Creek, CA, which runs through Markleville. As with Truckee, initial QPFs for Woodfords were underestimated. An update (Table 3) and (Table 3A) issued at the same time as the Truckee update performed better against actual precipitation.

NCEP guidance was helpful to the staff of NWSFO Reno in assembling local QPFs. NCEP issues daily 24 h QPFs valid through 12 UTC the next day. The initial forecast is issued at 06 UTC, while an update is issued at 10 UTC. These are shown in Figs. 7-13. (Fig. 7). The update raised the QPF for Truckee to 3-4 in, while QPF for Woodfords was unchanged (Fig. 8). Actual precipitation for 30 December was 1.35 in at Truckee and 1.21 in at Woodfords. The NCEP QPF overestimated precipitation amounts. It is important to note that observed precipitation is reported over the 24 h period beginning at 7AM LST (15 UTC) and this may cause some confusion in the verification of daily QPFs. Errors in NCEP QPFs can be attributed to the lack of local scale data availability.

Figure 9 shows the initial QPF forecast for 31 December, which was 3-4 in for Truckee and 2 in for Woodfords. The update (Fig. 10) dropped Truckee to 2-3 in and Woodfords to 1 in. This was the correct action as actual precipitation for 31 December was 1.03 in at Truckee and 0.21 in at Woodfords. QPF was again overestimated for 1 January 1997 (Fig. 11) and (Fig. 12). For 2 January, NCEP QPF was accurate as the initial forecast called for 3-4 in at Truckee (Fig. 13). Actual precipitation for this day was 4.65 in Woodfords QPF was underestimated somewhat at 1 in, compared to 2.90 in of actual rainfall.

NCEP also issues 48 h QPF products. These products generated a wider range of QPF values, but were fairly accurate. A summary of QPF activity for the period 30 December 1996 through 3 January 1997 is shown in Table 5. NWSFO Reno overestimated QPF at Woodfords and was within range for Truckee through this period. NCEP guidance overestimated QPF for Truckee. This overestimation was slightly worse with the 10 UTC updates. The 48 h forecasts were actually an improvement for this period. NCEP guidance was within range for Woodfords for both the initial and updated QPFs. It was overestimated somewhat in the 48 h QPF.


Conclusion

QPFs issued by NWSFO Reno were examined for their performance during a major flooding event. In developing local scale QPFs, forecasters at Reno utilized model- derived QPFs and real-time precipitation data obtained from automated gages and spotters. Incorporating a superior knowledge of synoptic and mesoscale weather patterns that effects the eastern Sierra Nevada Mountains and western Nevada, QPF forecasts issued from Reno verified very well. QPFs issued from NCEP were found to overestimate precipitation for this event. This would be expected in that NCEP does not have access to local scale data.

Information critical to forecasters when assembling QPFs includes freezing levels, snowmelt levels and real-time precipitation data. Automated gages must be constantly interrogated to provide the most recent precipitation data. Weather spotters are another good source of data. QPFs must always be monitored for accuracy and should be updated if conditions warrant; RFC streamflow models are highly dependant on accurate QPFs. These models are vital in issuing flood warnings to the public.

Gages must be maintained to perform accurately during times of heavy precipitation as this is the time that they are most needed. Improved gage networks would lead to more accurate QPFs. The NWS could then issue more timely warnings for flooding events and ultimately minimize the loss of lives and property.