Queries & Spatial Joins

The following situations are theoretical and intended as an educational exercise only. Information on current monitoring of Mount St. Helens can be found through the USGS.

Problem Statement

Queries and spatial joins can be essential tools for finding features of interest as well as examining spatial relationships. This analysis aims to apply queries and spatial joins to the following series of questions to begin to explore how some of the effects of the 1980 eruption of Mount St. Helens could look today.

Attribute Queries

• How many comparable eruptions to the 1980 Mount St. Helens eruption in terms of the Volcanic Explosivity Index have occurred in the United States in recent geologic history?

• How many volcanic vents in Washington are associated with Mount St. Helens?

Spatial Queries

• What areas could be within the Direct blast zone and Channelized blast zone, and what percent of it is on U.S. Forest service land?

• Which cities could be within a quiet zone in terms of not being able to immediately hear a lateral blast from an eruption?

Spatial Joins

• Which geologic units are each volcanic vent on Mount St. Helens formed within?

• What are the total lengths of U.S. Forest Service roads and trails in each potential blast zone that could be blocked or destroyed?

Combination of Queries and Joins

• What is the total population that could be most directly affected by lahar flows in the event of an eruption?

• How much of the combined blast and lahar hazard zones are developed in terms of land use?

Data

Mount St. Helens is located within the Gifford Pinchot National Forest in southwest Washington (Figure 1). Prior to the infamous 1980 eruption, the area was used for timber harvesting, mining, hydroelectric power generation, and recreational purposes. Although it was recognized as being an active volcano, the practices of monitoring and forecasting were relatively new to the Cascade Range of the Pacific Northwest (Wright et al. 2023). On May 18, 1980, a magnitude 5.1 earthquake triggered the most devastating volcanic disaster in the recorded history of the U.S. The northern flank collapsed into a massive debris avalanche, a lateral pressure blast expanded outwards for miles, pyroclastic explosions caused extensive damage, lahar flows resulted in downstream flooding, and ash spread eastward across the United States (Tilling et al. 1990). The lives of 57 people were lost in the eruption, and a total estimated $1.1 billion in damages were caused. Not only has the science behind volcano monitoring and forecasting improved in the subsequent decades, but also emergency management response and hazard planning. An integrated Volcanic Risk Management System has evolved for Mount St. Helens to facilitate partnerships and communications to more effectively prepare and respond to future volcanic activity (Wright et al. 2023).

Data for this project include features at the state, national, and global levels. The National Center for Environmental Information tracks worldwide significant volcanic eruptions today as well as estimates for past events based on historical accounts and data. U.S. data include National Park boundaries, National Forest Service roads and trails, and major U.S. cities. For the state of Washington, data include general land use information, major rivers, surface geological units, and the locations of volcanic vents. Descriptions of all data sources used can be found in Table 1.

Methods

The projected coordinate system used for this analysis is the NAD 1983 HARN State Plane Washington South FIPS 4602 (U.S. Feet). This system uses a Lambert Conformal Conic projection as it extends across southern Washington. Although the increase in accuracy from the HARN upgrade is likely not necessary for this analysis, the data obtained from Washington State were already utilizing it. 

Attribute Queries

• The Volcanic Explosivity Index (VEI) is a logarithmic scale used to describe volcanic eruptions based on magnitude and intensity. The 1980 eruption of Mount St. Helens was classified as having a VEI of five. The expression shown in Figure 2 was used to select all volcanic eruptions with a VEI of at least five that were within the United States and occurred relatively recently during the Holocene (defined as later than 0 CE). 

• Volcanic vents are features where lava flows onto the surface or pyroclastic material can erupt. The volcanic vents associated with Mount St. Helens were selected from across the state of Washington by the expression in Figure 3.

Methods

Spatial Queries

• The devastation from the lateral blast of the 1980 eruption was classified into two categories. The Direct blast zone, also known as the tree-removal zone, destroyed virtually everything for approximately eight miles from the peak. The Channelized blast zone, also known as the tree-down zone, flattened everything in its path for as far as 19 miles from the peak (Tilling et al. 1990). To recreate this area, two non-overlapping buffers were created around the peak at distances of eight and 19 miles (Figure 4). Next, the boundary of the Gifford Pinchot National Forest was clipped to the blast zone buffers to determine how much of this hazard area is U.S. Forest lands (Figure 5).

• The sound of the lateral blast from the 1980 eruption could be heard for hundreds of miles across the Pacific Northwest, yet was not reported as being heard in areas relatively nearby to the peak. This quiet zone, which extended at least as far as 50 miles to Portland, OR, was thought to be a result of the soundwaves responding differently to variations in atmospheric temperature and moisture (Tilling et al. 1990). To determine the nearby population that could be at risk of not hearing the initial eruption, major U.S. cities were selected that were within 50 miles of the peak (Figure 6).

Methods

Spatial Joins

• The ages of rock deposits of a volcano can indicate the frequency of activity and dormancy as well as the types of hazards it can produce. Understanding the past behavior of a volcano is essential in assessing future hazards. Information on the surrounding geology was joined to volcanic vents selected from the previous attribute query based on which map unit each vent was located within (Figure 7).

• Continuing from the spatial query regarding Gifford Pinchot National Forest, the total lengths of U.S. Forest Service roads and trails that could be affected by the blast zones were determined. The intersection of each feature with the blast zones was found to calculate the total lengths within each blast zone (Figures 8 & 9). 

Methods

Combination of Queries and Joins

• Lahars, also known as mudflows, are mixtures of volcanic debris and water that can be highly destructive on steep slopes following pyroclastic eruptions. Immediately following the 1980 eruption, lahars began flowing down stream channels from the peak and eventually discharged sediments into the Columbia River (Tilling et al. 1990). To determine where these lahars could occur, headwater streams were first selected within a distance from the peak (Figure 10). Visually tracing these streams to their confluence with the Columbia River led to the development of the expression in Figure 11 for selecting all of these streams at once. A 0.5 mile buffer was then placed on both sides of these stream lines to create a lahar channel hazard area (Figure 12). The population of each city within five miles of the lahar hazard areas were then summed based on a merge rule of the spatial join in Figure 13 to estimate the total affected population.

• A union overlay was first used to combine the lahar channels with the blast zone buffers to create an overall hazard area (Figure 14). The intersection of the general land use classes with the overall hazard area was then found (Figure 15). This intersection was then dissolved by each land use code and grouped together into the broader categories of residential, manufacturing, transportation, trade, services, recreational, agriculture, public forest, water, and undeveloped. Finally, the expression in Figure 16 was used to select all of the developed land use areas within the overall hazard area.

Assumptions

For finding recent significant volcanic eruptions, the year of 0 CE or later was chosen as events listed as BCE have more uncertainty associated with them. When determining the potential blast zones, the distances were set as uniform rings to match the distances that were observed in the 1980 eruption. Tilling et al. (1990) reported observations that the sound of the 1980 eruption was not heard approximately 50 miles away in Portland, OR. The distance for selecting cities that could be within this quiet zone was thus set exactly to 50 miles, even though if it occurred during an eruption again it would likely vary due to atmospheric conditions. For the intersections of the blast zones with the U.S. Forest Service roads and trails, the outputs were set to lines in order to determine the total lengths of the features within each zone. 

When finding the headwater streams, the search distance was set to 9 miles after several attempts to avoid selecting streams in watersheds that do not directly drain from the peak, such as the Coweeman River to the west. The line segments were also manually split at stream junctions on the Lewis, Cowlitz, and Columbia Rivers to avoid selecting upstream areas. The width of lahars can extend from within stream channel banks up to miles outside of them depending on the topography, eruption intensity, and amount of water present, so a value of 0.5 miles was chosen on either side as a general estimate. As the populations of the major cities were represented as points, a distance of 5 miles from the lahar flows was chosen to better account for the actual areas of the nearby cities. The 55 individual land use classes were grouped into ten categories to result in more readily interpretable results. Areas that were considered to be classified as developed included residential, transportation, recreational, manufacturing, services, and trade.

Results

Attribute Queries

• The returned list of significant volcanic eruptions in the U.S. is shown in Table 2. Only three unique volcanoes met the criteria, including Mount St. Helens. VEI values of 5 are described as being paroxysmal, such as both recent Mount St. Helens eruptions. The Churchill and Novarupta volcanoes in Alaska had VEI values of 6, which are described as being colossal. No new eruptions meeting the criteria have occurred since the 1980 Mount St. Helens eruption.

• Of the 508 known volcanic vents within Washington, 18 are associated with Mount St. Helens. These vents are displayed in Figure 18 as well as further described in Table 5. This excludes the volcanic vents of the nearby Mount Adams to the east.

Results

Spatial Queries

• The 8-mile Direct blast zone buffer as well as the 19-mile Channelized blast zone buffer are shown as non-overlapping rings in Figure 17. Their areas, as well as the total blast zone area, are given in Table 3. Approximately 58% of this blast zone is within the Gifford Pinchot National Forest. The land uses of other areas of the blast zone will be explored further below with Table 7.

• 19 major U.S. cities, defined as having populations of at least 10,000 per the 2020 U.S. Census, were found to be within 50 miles of the peak of Mount St. Helens (Table 4). If the conditions of the quiet zone were repeated with an eruption today, this indicates that over 1 million people across southern Washington and northern Oregon might not immediately be aware of the volcanic blast.

Results

Spatial Joins

• The 18 volcanic vents of Mount St. Helens were located within four geologic map units (Figure 18). Although they can be composed of different rocks, they are all associated with volcanic activity through lava flows on the surface or crystallizing magma beneath the Earth's surface (Table 5). They are also all deposited within the Quaternary period which has spanned from approximately 2.5 million years ago to the present.

• The lengths of U.S. Forest Service roads and trails that could potentially be closed or destroyed within the Gifford Pinchot National Forest by volcanic activity are given in Table 6 and displayed in Figure 17. There are overall more open roads that could be affected, primarily in the outer Channelized blast zone. There are less miles of trails affected, but their lengths are roughly equal within each blast zone.

Results

Combination of Queries and Joins

• The cities of St. Helens in Oregon and Kelso, Longview, and Ridgefield in Washington were all found to be approximately within 5 miles of the lahar hazard areas that flow from Mount St. Helens to the Columbia River to the southwest. Their individual 2020 populations are given in Table 4, and thus the total overall population that could be most directly impacted by lahar flows is 74,674.

• The areas of each land use category within the combined blast zone and lahar channel hazard areas are given in Table 7. Only 3% of the total hazard area of 1,245 square miles was considered developed, primarily along the lahar channels towards the cities along the Columbia River to the southwest (Figure 19). However, the destruction of trees and crops within the public forests and agricultural lands could still have devastating consequences on natural resources.

Conclusion

A primary limitation of this analysis is the assumption that a contemporary volcanic eruption of Mount St. Helens would be at the exact magnitude and intensity of the 1980 eruption. What was once a relatively natural mountain peak is now a horseshoe-shaped crater, meaning a massive landslide would not occur in the same way. The actual blast zones from the 1980 eruption were irregularly shaped and would vary today based on topography as well as the intensity of the eruption. The widths of any lahar flows would also vary depending on the amount of material present and the shapes of the stream channels. Also absent from this analysis is consideration of the glaciers present on Mount St. Helens, which could affect the amount of water in the system that could lead to lahar flows. Determining the total affected populations is limited by the use of point data for cities instead of areal units such as census tracts.

However, information obtained from these queries and spatial joins can still be useful towards hazard planning within the areas surrounding Mount St. Helens. From 2004 to 2008 a series of comparatively small volcanic events occurred on Mount St. Helens, although with a notable lack of explosive events (VEI of 2). The intensity of public interest in these events led to a number of communication challenges for scientists and government officials, but also to an increase in scientific research and monitoring as well as process improvements to risk management and emergency response (Wright et al. 2023).

There are only two other volcanoes with significant eruptions in the U.S. for which comparisons in magnitude and damages can be made. The 18 volcanic vents present are formed in four geologic units of the Quaternary period. Approximately 58% of a potential blast zone is within the boundaries of the Gifford Pinchot National Forest, and this blast could affect over 1,000 miles of roads and trails within the National Forest. 19 cities with a combined population of over 1 million people might not be able to hear the blast from an eruption, and a population of at least 74,000 might be impacted by lahar flows traveling down stream channels. Of the combined lahar and blast hazard area, only 3% of the land is developed but the effects on the 92% of land that is forest and agriculture could still be highly detrimental.

References

Tilling, R. I., Topinka, L. J., & Swanson, D. A. (1990). Eruptions of Mount St. Helens: Past, Present, and Future. U.S. Department of the Interior, Geological Survey.

Wright, H. M. N., Driedger, C. L., Pallister, J. S., Newhall, C. G., Clynne, M. A., & Ewert, J. W. (2023). Development of a volcanic risk management system at Mount St. Helens—1980 to present. Bulletin of Volcanology, 85(10), 53. https://doi.org/10.1007/s00445-023-01663-y