Landslides in Southeast Alaska: Precipitation as a Trigger
My capstone project for the MGIST program was in cooperation with NOAA’s National Weather Service Juneau office. It was an opportunity to combine past experience as a meteorologist with the skills learned throughout the MGIST program. The goal was to create a situational awareness tool to help forecasters maintain awareness of current precipitation accumulations across southeast Alaska. Precipitation has been shown to be a trigger of landslides, especially when higher amounts are seen within short periods of time (e.g. 3 hours of continued heavy rain has been known to trigger slides). The images in Figure 1 show damage from a fatal landslide in Sitka from 2015, a huge Glacier bay debris flow/landslide and the aftermath of the Lituya Bay mega-tsunami.
Glacier Bay landslide
Figure 1. The above images exemplify the damage that can happen due to landslides in southeast Alaska.
Analysis Procedures
Three web applications were developed to help forecasters stay aware of current precipitation conditions across southeast Alaska. These were an ESRI Story Map for education purposes, a web application that allows forecasters to query precipitation thresholds and a static map that shows current precipitation observations. Precipitation thresholds were determined for each of southeast Alaska’s 13 forecast zones using observed landslide locations linked with the weather at the time of the landslide. Figure 2 shows initiation locations for slides in Juneau, Sitka and Ketchikan with the observed 24-hr precipitation amounts.
Juneau slide locations
Ketchikan slide locations
Sitka slide locations
Figure 2. The maps above show known slide initiation points with the recorded 24 hour precipitation totals from the closest weather observing station.
To help classify precipitation accumulations into categories that forecasters can look at to decide on conditions that may produce slides, precipitation frequencies for 53 locations across southeast Alaska were examined. This information can be downloaded from NOAA’s precipitation frequency data server (PFDS) at https://hdsc.nws.noaa.gov/hdsc/pfds/. For each of the 53 stations, data is available in 1, 2, 5, 10, 25, 50, 100, 200, 500 and1000 year return intervals. For each return interval, precipitation durations for 5 min., 10 min., 15 min., 30 min., 60 min., 2 hour, 3 hour, 6 hour, 12 hour, 24 hour, 2 day, 3 day, 4 day, 7 day, 10 day, 20 day, 30 day, 45 day and 60 day are available.
Results
The results of the precipitation analysis indicated that the 10 year return intervals best match the observed precipitation that had occurred before observed slides. Using a File Geodatabase, an ESRI Story Map was created, along with a web mapping application and a static map of current observations categorized by year return interval. More information is available upon request.
Reflection
I was very excited to work with the NWS office as use of GIS is a skill that all forecasters utilize during every shift they work. Combining the spatial component of weather observations into a landslide risk statement could help provide advanced warnings of landslides to the citizens of southeast Alaska. The project was a culmination of the skills learned throughout the program and using those skills to solve a weather forecasting problem. One of the highlights of the trip was a personal visit to Juneau were I had the opportunity to discuss the project with those who would be using the information to produce a forecast. This was essential in determining the direction of the project. Two key elements that came out of the visit was that the spatial scale of the analysis and results should be the 13 forecast zones and no larger (large scale as in higher resolution) plus the PFDS information should be utilized during the analysis portion of the project. The latter realization was a result of attending a scientific talk put together by the Juneau National Weather Service, the US Forest Service and Juneau’s Emergency Manager. Having the opportunity to spend several days listening to what the NWS would find the most useful and getting their insight into the project was an invaluable experience.
Carrie L. Suffern 