Overview:

The winter storm of January 23/24, 1999 in Southeast Alaska was an event that was well covered by the National Weather Service, Weather Forecast Office (WFO), Juneau, Alaska. Parts of the northern and central panhandle of Southeast Alaska received ten to fifteen inches of snow in less than 24 hours. Most notable of this event was the long lead time provided to the public by WFO Juneau. A winter storm watch for heavy snow was issued Thursday afternoon, January 21, for the following Saturday, January 23. The watch was issued to cover the northeast Gulf Coast, and northern and central Southeast Alaska, which is most of WFO Juneau's forecast area of responsibility. The purpose of this paper is to document the event, and discuss what model data gave the duty forecaster(s) the confidence to issue the winter storm watch 48 hours in advance of the event.

Prelude:

Synoptic Discussion:
On Thursday January 21, a split flow existed at the mid and upper levels of the atmosphere over the Northern Pacific and Gulf of Alaska, from 700 millibars (mb) (roughly 10,000 ft) up through 300 mb (roughly 30,000 ft). This is illustrated in figures 1a - b - c, showing the 12Z (3am Alaska Standard Time) 1/21/99 300, 500 (about 18,000 ft), and 700 mb analyses. A mean ridge had existed for several days previously over northwest Alaska in the weaker northern flow branch of the jet stream. The stronger southern branch contained systems that were forming in the North Pacific off the coast of Siberia and moving east in the 45-53 degrees north latitude band. GOES IR imagery at 2330Z (230pm AST) on Thursday Jan. 21, 1999 (Photo at left) showed a large moisture band/occluded front associated with a still-deepening low just off the Siberian coast near the Kamchatka peninsula. The appearance of this frontal band, with its long moisture fetch, extending back to the subtropics, combined with the MRF/AVN model forecasts for the next three days, gave the duty forecaster sufficient confidence to issue the winter storm watch at 400 pm Thursday, January 21 for the following Saturday, January 23. The decision was based on the fact that some heavy snowfall events in Southeast Alaska are generated when moisture ahead of Pacific occluded fronts (the warm conveyor belt) "over-runs" an established low-level arctic air-mass over the inner channels and immediate coastline. This arctic air-mass is maintained by the outflow of cold arctic air from the interior of Alaska and the Yukon territory. Lifting of the maritime air-mass combined with the sub-freezing lower- layer temperatures combine to produce the event. A sub-freezing low-level layer or arctic air-mass, had evolved over Southeast Alaska for several days previous to January 21. Weak systems with ample clearing behind each one, moved over Northwest Alaska and into the Yukon. Dynamically-forced subsidence (sinking air) combined with strong radiational cooling caused a 1040+ mb "arctic high" to form in the Alaska interior by Thursday, January 21. The cold air was deep enough to flow through the coastal mountain gaps and over the passes into Southeast Alaska. The 18Z (9am AST) Jan. 21 Alaska surface analysis shows this (figure 3), the 1041 mb high in the Alaska/Yukon interior with an arctic frontal boundary in the Gulf of Alaska and North Pacific. The deepening 975 mb parent low of the system of concern is shown at 53 degrees north, 175 degrees east, as well.

The Main Event:

By 12Z (3am AST) Saturday, January 23, the frontal zone had moved east, as the parent upper-level low moved northeast into the Bering Sea near St. Paul Island. The upper-level ridge which had been in place over Western Alaska and the Bering Sea had crossed the Gulf of Alaska and was now aligned southwest to northeast over the Southeast Alaska panhandle. The surface arctic boundary was still in place just offshore of the panhandle as cold arctic air continued to flow out of the interior from the arctic high. The arctic high, in turn, began sliding southeast in response to the upper-level changes. Figure 4 shows the Alaska surface analysis at this time. Note the cold air in place at the surface, temperatures of -19F at Whitehorse, YT, and Dease Lake, BC, and 17F at Juneau. Significant snowfall began in Yakutat after 13Z, as seen in table 1, surface observations. Significant snowfall did not begin in Juneau until 22Z (1pm AST) however (table 2).

The upper-level and surface analyses for 00Z Sunday January 24, or 3pm AST, show the event at its peak. The surface analysis (figure 5a ) shows the main surface low with two centers in Western Alaska with the arctic boundary just near Yakutat but still south of Juneau, while the occluded front stretched back into the central Gulf of Alaska and North Pacific. A strong, saturated southwest flow existed at 500 mb through 850 mb (roughly 6,000 ft) (figs. 5b - c) with warming temperatures, a typical over-running scenario, while surface temperatures at Juneau were still in the 20's (table 2). Yakutat, because of its open exposure to the southwest and the Pacific Ocean, warmed to 34 degrees at 00Z, and the snowfall changed to rain. Low-elevation sites in the central and northern panhandle of Southeast Alaska at this time, from Petersburg and Wrangell north to Skagway and Haines, were all reporting snow, and continued to do so for many more hours as the event progressed. One other feature worthy of mention at this time is the 300mb analysis, figure 5d. A strong southwest jet with a maximum wind speed core in excess of 130 knots is in the North Pacific, aimed at the panhandle. The northern panhandle and eastern gulf coast are on the left-front quadrant of this feature, and remained so through 12Z (3am AST) Sunday, as can be seen in figure 5e. This is a location of favorable dynamics for large-scale lifting and so was certainly a factor in the duration and amount of snowfall received in the northern panhandle.

A GOES IR image from 21Z (Noon AST) Saturday, January 23 (Photo at right) shows the main frontal band moving into the panhandle, at which time the snow in Juneau was just about to begin (see table 2).

Wind profiler time-series data from Lemon Creek, about three miles east of the Juneau airport, proved to be very informative for the duration of this event. The profiler data clearly showed the southeast winds associated with the colder low-level arctic air-mass, up to three to four thousand feet above mean sea level ( m.s.l). Winds veer to the southwest between four and seven thousand feet m.s.l., the lower-levels of the over-running associated with the frontal band.

Snowfall amounts for this event (noon Sat. 1/23 to noon Sun. 1/24) were as follows.

In the Juneau vicinity:

Other reports from sites in Southeast Alaska include:

Model Performance:

The goal of this section is to show what forecast model data influenced the duty forecaster(s) to issue a winter storm watch with 48 hours of lead time. While on duty before and during this event, I was working the public shift, and did not issue the preliminary watch, Thursday afternoon, January 21. A combination of model forecasts and interpretation of satellite imagery were used in the decision to issue the watch, however. The AVN forecast model, which runs out through 72 hours, is one of a suite of computer-generated weather forecasting models produced twice daily by the National Center for Environmental Prediction (NCEP), in Washington DC. The other models used most frequently by the NWS are the short-range (0-48 hours) ETA and NGM, and the longer range MRF (0-240 hours). All these models use input data from many sources, satellites, surface land and sea-based observations, balloon soundings, and airplane reports. This data is input to these models which simulate the current weather patterns over North America and the entire Earth, then generate forecasts based on the current pattern, all using very complex physics to simulate the many processes at work.

The AVN 500mb height forecasts for 60 and 72 hours were a little slow on the timing and slightly weak on the strength of the southwest flow at 500mb. The 60 and 72 hour AVN surface and 1000-500mb thickness forecast charts compared favorably with the actual analyses for the same time period however. The model forecasts clearly showed a strong warm advection pattern developing, one indication of the significant over-running precipitation potential, though it was also a little slow on the timing of the features and over-forecast the strength of the surface high in the North Pacific, south of Alaska. The same charts for the same time period from the 00Z (3pm AST) Jan. 21 run of the MRF also verified fairly well with the analyses at 00Z and 12Z Jan. 24. The 00Z Jan. 21 MRF model run was also a little slow on the timing of the overall pattern progression, however, and also over-forecast the strength of the 500mb ridge over the North Pacific, and the corresponding surface high.

The ETA model run from 12Z (3am AST) Friday, January 22 proved to be fairly accurate in the timing and strength of the system, and gave the duty forecasters that day continued confidence in maintaining the heavy snow watch for the following day. The ETA 36 and 48 hour 850 mb height and temperature forecast charts clearly showed a strong southwest flow with warmer maritime air moving into the much colder arctic air in place, and compared favorably with the 850mb analyses for 00Z and 12Z January 24. The ETA 36 hour and 48 hour 850 mb temperature forecasts in fact were highly accurate in this case. This is important because the 850mb (roughly 6,000ft) temperature is a critical value used by forecasters in WFO Juneau to assess snowfall potential; 850mb temperatures greater than -4C generally are associated with rain at sea level sites in Southeast Alaska. Other ETA forecast charts from the same 12Z Friday January 22 run for the same time period were equally as accurate and detailed in forecasting this event. The 36 and 48 hour ETA 500 mb height forecasts showed the strong southwest flow developing , and the ETA 36 and 48 hour 700mb height/vertical velocity forecasts showed the strong southwest flow as well as significant upward vertical motion moving into the panhandle. The ETA 36 and 48 hour Quantitative Precipitation Forecast/1000-500 mb relative humidity forecast charts, and the corresponding 1000-500 mb thickness and surface pressure forecast charts forecasted a significant moisture field and precipitation amounts for the panhandle during the evening of Saturday, January 23 through the early morning of Sunday January 24, along with strong warm-advection in the 1000-500 mb thickness field, key signatures of a warm over-running snowfall event. While not totally accurate in all details, mainly in the structure and strength of the surface lows, this 12Z Friday, Jan. 22 ETA model run was on track with the timing and strength of this system. The Nested Grid Model (NGM) also had a good overall handle on this system in the 36 and 48 hour forecast time period; it agreed in general with the ETA model forecast information, specifically concerning the depth of the moisture field and the 850 mb temperatures.

A GOES IR image from 18Z (9am AST) January 22 (Photo at left), shows the frontal band still maintaining a strong subtropical moisture feed, holding together, and movement of the parent surface and upper-level low in the model-forecasted northeasterly direction. This, combined with the continuity of the models in forecasting this event, maintained the forecaster's confidence of its occurrence such that the Winter Storm Watch for heavy snowfall was upgraded to a Winter Storm Warning at 1300Z (4am AST), January 23. The warning of course, verified in the area of its coverage, the eastern Gulf Coast and northern Southeast Alaska.

Summary:

The heavy snowfall event of January 23-24, 1999 over Southeast Alaska was a typical warm-advection/over-running precipitation case, where low-level arctic air provided a net sloping surface for ascent of warmer moist air above 4000 feet m.s.l., while the low-level cold air layer was able to maintain a depth necessary to keep precipitation in the form of snow. MRF and AVN model runs at 60-84 hours along with GOES IR imagery analysis gave the duty forecaster on Thursday, January 21 sufficient confidence to issue a winter storm watch for heavy snowfall at 400 pm, 48 hours in advance of the event. Subsequent model runs, especially the ETA model, continued to show a significant snowfall event for Saturday evening/Sunday morning 1/23-1/24 developing, and helped maintain forecasters' confidence in the event. Overall this was an excellent example of the National Weather Service's capability of providing valuable and timely weather information to the public well in advance of a potentially threatening weather event. This example also clearly illustrates one of the categories of weather systems that generate heavy snowfall in southeast Alaska.

PAYA     SKY CONDITIONS       VSBY WX      SLP     TT  DP   WIND    ALT     APP
231200      OVC027              10 S-      275     21 18    0000    034
231253      OVC025               3 S-F     268     23 20    0000    033     -23
231324      OVC015               2 S-F     265     21 19    0803    031
231353      OVC013           1 1/2 S-F     255     23 21    0805    029     -31
231453      BKN017 OVC037    1 3/4 S-F     246     25 24    0906    026     -26
231553      BKN013 OVC022    1 1/2 S-F     232     25 24    0906    022     -36
231653      VV004              3/4 S-F     215     27 26    0909    017     -40
231753      BKN004 OVC011      3/4 S-F     204     29 28    1109    014     -42
231853      0VC002             1/4 S+ZF    195     30 29    1109    011     -37
231953      VV002              1/4 S+ZF    180     31 29    1213    007     -35
232053      BKN004 OVC013      1/4 SF      161     32 30    1313    002     -43
232153      0VC007               2 S-F     143     34 32    1114    996     -52
232253      BKN006  OVC015   2 1/2 R-S     132     34 34    0804    993     -48
232353      BKN013 OVC018        4 R-S     127     34 33    0304    991       
240053      BKN023 0VC041       10 R-      120     33 33    1005    989     -23
240153      BKN013 OVC021        4 RF      116     34 33    1206    988     -16
240253      OVC004               4 R-F     108     33 33    1105    986     -19
240353      0VC004               4 RF      101     33 33    0604    984     -19
240453      SCT004 OVC012       10 R-      096     33 33    1108    982     -20
240553      BKN024 OVC032        6 R-F     091     34 33    1107    981     -17
240653      OVC015              10         086     33 32    1007    979     -15
240753      BKN010 OVC015    1 1/4 S-F     080     33 32    1205    977     -16 
240853      BKN007 OVC018        2 S-F     075     33 32    1304    976     -20
240953      BKN006 OVC014    1 1/4 S-F     071     32 32    0000    975     -15
241053      BKN015 OVC032        6 R-S     067     33 32    1405    973     -13
241153      BKN005 OVC014        1 S-F     066     32 32    1504    973     -09
241253      BKN009 OVC030        3 RS-F    064     33 32    1305    972     -11
241353      BKN045 OVC060        6 F       060     33 32    1204    971     -07
241453      OVC004               3 F       059     32 32    1104    971     -07
241553      OVC002           1 3/4 F       057     32 32    1404    970     -07
241653      OVC002               2 F       057     32 32    1205    970     -03
241753      OVC002               4 F       055     33 32    1407    970     -04
241853      OVC004               3 F       057     33 33    1104    971      00
241953      OVC004               3 RS-F    057     33 32    1207    971      00

Table 2, Juneau Airport Surface Observations, 19Z Jan. 23 to 20Z Jan. 24

PAJN     SKY CONDITIONS     VSBY WX         SLP     TT DP   WIND   ALT      APP
231753      OVC041            10            274     20  9   0713    034     -18
231853      SCT034 OVC041      9 S-         270     20 11   0917    033     -16
231953      SCT034 OVC041     10            264     21 12   0815    031     -18
232053      BKN028 OVC041      7            257     23 13   0914    029     -17
232153      BKN015 OVC025      5 S-         246     24 15   0815    026     -24
232253      VV014              3 S-F        238     23 19   1017    024     -26
232353      VV014              3 S-F        229     24 21   0916    021     -26
240053      VV014              2 S-F        222     25 23   1016    019     -24
240153      VV007          1 1/4 S-F        212     26 25   1120    016     -26
240253      VV008              1 S-F        200     27 25   1122    012     -29
240353      VV006            3/4 S-F        191     27 26   1119    010     -31
240453      VV006            3/4 S-F        180     27 27   1118    006     -32
240553      BKN005 OVC010      1 S-F        168     28 27   1223    003     -32
240653      OVC010         1 3/4 S-F        156     28 27   1120    999     -35
240753      BKN010 OVC015  1 3/4 S-F        143     29 28   1018    995     -37
240853      BKN007 OVC010  1 1/4 S-F        131     29 28   1116    992     -37 
240953      VV006              1 S-F        120     30 29   1117    989     -36
241053      VV006              1 S-F        109     30 29   1014    985     -34
241153      OVC009             2 S-F        099     30 30   1114    982     -32
241253      BKN007 OVC012  1 1/2 S-F        089     31 30   1012    979     -31
241353      VV006              1 S-F        082     31 31   1011    978     -27
241453      VV005              1 S-F        078     31 31   0910    976     -21
241553      VV007          1 1/2 S-F        073     31 31   1111    975     -16
241653      VV007          1 1/2 S-F        075     32 31   0809    975     -07
241753      VV010              2 S-F        081     32 31   0906    977     +01
241853      VV010          1 1/2 S-F        085     32 31   0905    978     +12
241953      VV013              2 S-F        088     32 32   0806    979     +13

---

References

1. Bader, M.J., Forbes, G.S., Grant, J.R., Lilley, R.B.E., and Waters, A.J, 1995: Images in Weather Forecasting, A Practical Guide for Interpreting Satellite and Radar Imagery. Cambridge University Press., 302-330.

2. Colman, Bradley, 1986: The Winter Climate of Juneau: A Mean of Contrasting Extremes. National Weather Digest, Volume 11, No. 2.

3. Kanan, R.A., 1979: The Arctic Front and Aviation Weather in Southeast Alaska During the Winter, Techical Information Paper No. 1. NOAA/National Weather Service/Alaska Region.

4. Tschantz, Bob, 1995: Winter Storm of December 4th, 1995. Local Study, National Weather Service Forecast Office, Juneau, AK.

---

Return to Case Studies home page