A Modeled Innundation and Water Level Storm Case:
The October 1963 Storm in Barrow Alaska
Introduction
The NCSU (North Carolina State University) CAPES model was used to model innundation and water levels for the 3 October 1963 storm in Barrow Alaska. The model was run for the coastline area starting southwest of Barrow up to and including the spit and Point Barrow (Figure 1). Model output was hourly starting at 1400 AST (Alaska Standard Time) on 2 October 1963 and running until 4 October 1963 at 1300 AST. Model input is as follows:
An important note, the topography and bathymetry used by the model is different than the topography and bathymetry in 1963. Be mindful of these differences, model output must be interpreted accordingly. For instance, in 1963 beach dredging was occuring, thus effecting beach topography, which will then effect timing and extent of flooding.
A full description of the storm can be found in the 1967 Hume and Schalk paper here.
Figure 1. Topography and bathymetry used for model input. Red box shows extent of the modeling domain. Click on this figure for a larger view and printable jpeg.
Innundation Composite
All 47 time steps were compiled into one layer, a composite, which displays any cell that was touched by flood waters. Innundation was derived solely by winds, topography, bathymetry and storm surge. Wave height is not included in this composite map, resulting in differences between the Hume and Schalk flood debris line and the innundation modeling results.
Figure 2. Note modeled innundation covered most of the spit, which coincides with the October event. Innundation follows the Hume and Schalk flood debris line near the Dew Line station and NARL (Navy Arctic Research Laboratory). Topography has changed between 1963 and 2002 at the landfill, at Isatquaq Lagoon (the bridge road is blocking flooding into the fresh water source), and the Dredge Harbor. Note there is no flooding over the beach berm near NARL. Flooding near the town of Barrow and Browerville stays close to the shoreline because of higher topography in that area.
Modeled Innundation and Drying
The following four pages of figures illustrate the progression of the CAPES modeled innundation overlaid on the 2002 Quickbird satellite image. Innundation on land along the Chukchi coastline is shown in black. Drying on the water in the Elson Lagoon is also shown in black. Model output times in GMT and Alaska Standard Time (AST) are shown at the top of each time step plot. The red line shows the Hume and Schalk flood debris extent from maximum flooding. This line was mapped after the storm.
Click on each set of 6 figures below for an 8.5" x 11" printable higher resolution copy.
Figure 3. CAPES modeled innundation for the duration of the October 1963 storm overlaid on the 2002 Quickbird satellite image. Note areas in Barrow and Browerville weren't flooded as extensively as NARL because there is higher tundra behind the villages.
Comparison of Modeled Innundation and Water Level to October 1963 Storm Event
Important Note: Water levels shown in maps below are modeled height above sea level, not height above ground elevation. To get actual flooding elevation above ground elevation the ground elevation must be subtracted from the numbers shown in the maps.
The next set of maps concentrate on modeled storm innundation and water elevation (or height/surge) for the NARL area. The most comprehensive storm observations were recorded by the Arctic Research Laboratory (ARL) Progress Report for October 1963 (pp. 12-17). The ARL acronym has since been changed to NARL. The maps below are a direct comparison of the modelled output with the text descriptions recorded at ARL.
Click on each map below for an 8.5" x 11" printable higher resolution copy.
Each figure below contains the following:
text taken directly from the ARL report with time stamp (AST) on 3 October 1963,
maps showing model innundation and surge for the corresponding time stamp (each map contains a time step in GMT and AST units), model output is overlaid on top of the 2002 Quickbird satellite image.
Notable geographic features are highlighted on the map,
Green shading in water areas inside Elson Lagoon are areas of drying. Final displays presented to the Barrow community will not include drying. Drying is defined as water areas that recede to a depth of .5 meters. In this case water is pushed away from the land by the force of the wind.
Figure 4. Modeled surge is rising in the Chukchi to match text description.
Figure 5. Modeled surge is dropping in Elson Lagoon and rising in the Chukchi. Modeled output is creeping onshore at NARL, but no water is crossing the road near the airport and filling Fresh Lake. This could be due to differences in beach topography between that used for modeling and what actually existed in 1963.
Figure 6. Surge is continuing to rise. Water is slowing moving towards camp, covering some of the road and some of the runway. The modeled innundation is moving more slowly than the actual storm, Fresh Lake contains no flood water yet. Surge depth on land varies from 0.7 - 1.1 meters.
Figure 7. A closer view of the same time period. Note, as the innundation modelling progresses through time, the runway is never really covered with flood waters. This could be due to a difference in topography and bathymetry from 2002 and 1963.
Figure 8. The innundation model doesn't show any water entering camp during this time period. There is some water touching the first row of buildings along the shoreline. By 1300 AST modeling shows that Fresh Lake is rapidly filling with floodwaters.
Figure 9. By now the innundation modeling shows Fresh Lake as mostly flooded. Buildings referred to in the text are not on this map. I need to work with a local to find them, if they still exist. Most of the main street into camp is flooded by the model. It looks like the modelled floodwaters are following the road as noted in the text description. Modelled water depth is too deep at this point, but it does seem to be in the right place.
Figure 10. Peak modeled innundation seems to be lagging behind actual storm occurances, but modeled flooding is quite extensive at this time step. Water levels are too high. Note the ARL building is always above the floodwaters, as actually happened during the storm event. The topography used to model the event did not include building elevations, just land elevations. The 1600 AST map shows the measurement noted in the text, "distance from camp to the airport runway".
Figure 11. Water depths are displayed on this plot. Note Elson Lagoon elevations are lower than those in and around NARL. Modeled water levels are still high at 1700 AST. Modeled water levels don't begin to drop until later.
Figure 12. By 1800 AST modeled water levels begin to drop as water equalizes with Elson Lagoon. During the actual event, storm water created a breech into Elson Lagoon (location shown on 1900 AST map) thus allowing flood waters to drain into Elson Lagoon. This could explain some of the discrepancy between modeled water levels and actual water levels.
Figure 13. Modeled water levels are still high the next day, 4 October at 0300 AST.
Figure 14. Modeled water levels rapidly recede on 4 October at 0400 AST.
Figure 15. Modeled water levels remain at this level for the duration of the model run, 4 October 0600 AST until 4 October 1300 AST.