{"id":266,"date":"2020-07-27T18:42:25","date_gmt":"2020-07-27T22:42:25","guid":{"rendered":"https:\/\/umaine.edu\/watershedresearch\/?page_id=266"},"modified":"2025-07-10T10:30:10","modified_gmt":"2025-07-10T14:30:10","slug":"vulnerability-analysis","status":"publish","type":"page","link":"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/","title":{"rendered":"Vulnerability Analysis"},"content":{"rendered":"

In order to effectively fulfill the mission of managing the health of Maine\u2019s shellfisheries, the Maine Department of Marine Resources requires adequate data and support tools to make scientifically-informed decisions about temporary closures in response to harmful coastal bacteria pollution events. To facilitate data-driven management decisions surrounding closure and re-opening of shellfish flats, School of Earth and Climate Sciences PhD student Bea Van Dam<\/a> and her advisor Sean Smith<\/a> have undertaken a full-coast bacteria pollution vulnerability analysis, using spatial data to investigate how proxy metrics related to bacteria sources, delivery from the landscape to coastal waters, and residence time in estuaries and embayments interplay to affect vulnerability to bacteria pollution in different coastal settings.<\/p>\n

<\/h3>\n

Project Background<\/h3>\n
\"Image<\/a>
Figure 1. A Source-Delivery-Residence Time ternary diagram provides a useful conceptual framework for understanding drivers of bacteria pollution vulnerability in varied coastal settings.<\/figcaption><\/figure>\n

The Maine Department of Marine Resources (DMR) is tasked with ensuring the safety of the state\u2019s valuable and extensive estuarine shellfisheries against such health threats as fecal coliform bacteria pollution. These bacteria and other pollutants are washed from the landscape during precipitation events and carried to via stream networks and surface flow to coastal waters, where they accumulate in clams and other filter feeders and can cause sickness in consumers of harvested shellfish. To mitigate these threats, shellfish managers temporarily close mudflats after large rain events, but this in turn negatively impacts the livelihoods of harvesters and their communities. In order to keep temporary closure lengths to the minimum required to ensure consumer safety, it is vital for shellfish managers to understand how various coastal settings differ in their vulnerability to bacteria pollution events, which is driven by differences in bacteria sources<\/em> on the landscape, delivery<\/em> from landscape to coastal waters (i.e. runoff generation and drainage networks), and residence time<\/em> within the estuaries before flushing out to sea (Figure 1).<\/p>\n

\"Map<\/a>
Figure 2. Margin Watershed Areas (brown) comprise much of the immediate coastal landscape, falling between non-tidal watersheds (yellow) and Gulf of Maine waters. Mount Desert Island, ME.<\/figcaption><\/figure>\n

Prior research approached this question by assessing source-, delivery-, and residence time-related proxy metrics for 535 stream and river watersheds draining to Maine\u2019s tidal coastline and finding clusters of watersheds with similar characteristics. This limited set of clusters, or \u201clandscape response units,\u201d allowed DMR to leverage limited monitoring capacity by applying knowledge gained from well-monitored locations to make informed management decisions in other harvesting areas within the same cluster.<\/p>\n

Certain data gaps were identified during the course of the initial proof-of-concept research, including the omission of much of the near-shore landscape from the analysis. The deltas of land between non-tidal watershed outlets, which we are referring to as \u201cMargin Watershed Areas\u201d (MWAs or \u201cmargins\u201d), do not support perennial stream networks but instead contribute runoff directly as overland flow (Figure 2). These areas comprise a total of ~1,800 km2 <\/sup>of Maine\u2019s coastal landscape and include parts of some of the state\u2019s most densely populated cities and towns, including the entirety of the Portland peninsula. This project began with the specific goal of incorporating these margin areas into an expanded vulnerability analysis, where the unit of analysis also shifts from the non-tidal watershed (river mouth) to the entire coupled landscape-estuary system (\u201cestuary unit\u201d).<\/p>\n

<\/h3>\n

Data Assembly<\/h3>\n

The first stage of this project required the delineation of surface flow paths for the length of Maine\u2019s coastline in order to incorporate the margin areas with the original 535 non-tidal watershed polygons and update the borders of those polygons where new elevation data had become available. Delineation occurred over a total of 5,236 km2<\/sup> of coastal land area and involved the placement of over 30,000 culvert lines to ensure proper flow routing across the 2 m resolution LiDAR bare earth digital elevation models (DEMs). This stage resulted in the creation of 12 output data layers (hydro-enforced bare earth and hillshade DEMs at 2 m and 1\/3 arcsecond resolutions, culvert polylines, flow direction and flow accumulation rasters at 2 m and 1\/3 arcsecond resolutions, and stream polylines based on 0.05, 0.20, and 0.30 km2<\/sup> area to channel initiation) to feed into both the vulnerability analysis and any future modeling needs.<\/p>\n

\"Map<\/a>
Figure 3. Non-tidal watersheds polygons (yellow) and gridded margin cell polygons (brown). Somes Harbor, Mount Desert Island.<\/figcaption><\/figure>\n

These data were used in conjunction with National Hydrography Dataset (NHD) stream lines to determine the placement of stream and river outlets along the length of the coast. Outlets were placed at the closest flow accumulation line to their \u201cofficial\u201d NHD counterparts. Due to the densification of the NHD stream layer in the intervening years, the number of non-tidal watershed polygons increased from 535 to 2,196, with the additional watersheds falling within area that had originally been considered margins. The remaining 1,800 km2<\/sup> of margin area was split into a grid of 30 m x 30 m polygon cells to accommodate calculation of proxy metrics (Figure 3).<\/p>\n

In the second stage of data assembly, existing spatial data layers were overlaid on the landscape polygons to calculate metrics related to bacteria sources, including land cover, population \/ density, and point source sites such as overboard discharge and combined sewer overflow sites; and delivery, including soil types and storage capacity, surface slope, and natural and engineered drainage density.\u00a0 Where appropriate, metrics were additionally broken into whole-watershed and margin-only subcategories.\u00a0 When combined with in-estuary residence time metrics (next section), each delineated estuary unit is described by a total of 107 columns of metric data.<\/p>\n

<\/h3>\n

Estuary Delineation GIS Tool<\/h3>\n
\"Map<\/a>
Figure 4. An “estuary unit” (bold outline) comprised of non-tidal watersheds, margin watershed areas, and an embayment, delineated from the its outlet line (yellow) at the mouth of Somes Sound, Mount Desert Island.<\/figcaption><\/figure>\n

Rather than provide DMR with a static map of a limited number of delineated estuaries, the decision was made to build a flexible tool that would allow users to draw an outlet line across any coastal embayment or sub-embayment and see source, delivery, and residence time metric data for that coupled land-sea unit.\u00a0 The ArcGIS tool, written in Python, incorporates the previously developed flow direction and flow accumulation data layers and uses those to delineate the contributing landscape and embayment area above each outlet line.\u00a0 All non-tidal watershed polygons and margin cells that fall within this area are selected, and their source and delivery time metrics are aggregated into landscape-wide totals.\u00a0 Next, the delineated embayment is used to calculate residence time related metrics, including embayment circularity and the ratio of landscape runoff volume to embayment volume for a 2\u201d rainfall event.\u00a0 The tool returns a polygon and metric attribute table for each estuary unit, and since the clustering exercise (next section) it has been updated to also calculate cluster membership for each newly delineated unit.<\/p>\n

Clustering and Vulnerability Analysis<\/h3>\n

500 tidal embayments and sub-embayments along the length of Maine\u2019s coast were delineated with the ArcGIS tool to feed into a cluster analysis, to find \u201ctypes\u201d of coastal settings. It is expected that members of the same cluster will exhibit broadly similar landscape and estuarine responses to precipitation events, allowing DMR to transfer knowledge and management strategies from well-monitored sites to others in the same cluster. For the preliminary cluster analysis, the data were standardized (each column centered by subtracting the mean and scaled to unit variance), then a principal component analysis was performed to reduce the dimensionality of the dataset from 107 metrics to 18 principal components, which explained just over 80% of variance in the data. Clustering was then performed on the principal component dataset using a Gaussian mixture model (GMM) with full variance. Since the GMM requires the user to specify the number of clusters, testing was done for k=4 through k=15 with 100,000 initializations for each k , and 8 was found to be the optimum number based on the Bayesian Information Criterion. The optimum model was then integrated back into the GIS tool to allow newly delineated estuary units to be assigned to a cluster. Explorations of more sophisticated clustering methods for the data are ongoing.<\/p>\n

\"\"
Figure 5. (left) Delineated embayment units colored by cluster, Mount Desert Island, ME. (right) Box plots show the distribution of bacteria counts for the eight preliminary clusters.<\/figcaption><\/figure>\n

480 of the 500 delineated embayments contain one or more DMR bacteria monitoring sites. Data from these sites will be used to perform a vulnerability analysis on the final clusters to determine whether certain types of coastal settings are more vulnerable to high bacteria loads than others and assess the tightness of the spread within each cluster, which will inform DMR management approaches.\u00a0 Early analysis of bacteria loads in the preliminary clusters indicates there may be some measure of equifinality at play \u2013 some clusters may have roughly similar median bacteria levels but for different reasons, for example, high source loads in one cluster vs high residence times in another, which would require different management responses (Figure 5).\u00a0 A full vulnerability analysis will be performed once clustering is finalized.<\/p>\n

\"Conference<\/a>
Poster detailing high-resolution elevation data preparation in locations adjacent to tidal estuaries along the coast of Maine. Presented at 2019 Maine Sustainability & Water Conference<\/figcaption><\/figure>\n
\"\"<\/a>
Poster summarizing the Estuary Builder spatial data tool designed to harvest land-sea connection attributes important to coastal management decision-making. First place award in the 2021 Maine Sustainability and Water Conference graduate student poster competition (Bea Van Dam).<\/figcaption><\/figure>\n","protected":false},"excerpt":{"rendered":"

In order to effectively fulfill the mission of managing the health of Maine\u2019s shellfisheries, the Maine Department of Marine Resources requires adequate data and support tools to make scientifically-informed decisions about temporary closures in response to harmful coastal bacteria pollution events. To facilitate data-driven management decisions surrounding closure and re-opening of shellfish flats, School of […]<\/p>\n","protected":false},"author":1767,"featured_media":0,"parent":1862,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_kad_blocks_custom_css":"","_kad_blocks_head_custom_js":"","_kad_blocks_body_custom_js":"","_kad_blocks_footer_custom_js":"","_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"footnotes":""},"class_list":["post-266","page","type-page","status-publish","hentry"],"yoast_head":"\nVulnerability Analysis - Watershed Process and Estuary Sustainability Research Group<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Vulnerability Analysis - Watershed Process and Estuary Sustainability Research Group\" \/>\n<meta property=\"og:description\" content=\"In order to effectively fulfill the mission of managing the health of Maine\u2019s shellfisheries, the Maine Department of Marine Resources requires adequate data and support tools to make scientifically-informed decisions about temporary closures in response to harmful coastal bacteria pollution events. To facilitate data-driven management decisions surrounding closure and re-opening of shellfish flats, School of […]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/\" \/>\n<meta property=\"og:site_name\" content=\"Watershed Process and Estuary Sustainability Research Group\" \/>\n<meta property=\"article:modified_time\" content=\"2025-07-10T14:30:10+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/umaine.edu\/watershedresearch\/wp-content\/uploads\/sites\/554\/2020\/09\/SDRT_diagram-e1599860134363.png\" \/>\n\t<meta property=\"og:image:width\" content=\"284\" \/>\n\t<meta property=\"og:image:height\" content=\"265\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data1\" content=\"9 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/\",\"url\":\"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/\",\"name\":\"Vulnerability Analysis - Watershed Process and Estuary Sustainability Research Group\",\"isPartOf\":{\"@id\":\"https:\/\/umaine.edu\/watershedresearch\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/umaine.edu\/watershedresearch\/wp-content\/uploads\/sites\/554\/2020\/09\/SDRT_diagram-e1599860134363.png\",\"datePublished\":\"2020-07-27T22:42:25+00:00\",\"dateModified\":\"2025-07-10T14:30:10+00:00\",\"breadcrumb\":{\"@id\":\"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/#primaryimage\",\"url\":\"https:\/\/umaine.edu\/watershedresearch\/wp-content\/uploads\/sites\/554\/2020\/09\/SDRT_diagram-e1599860134363.png\",\"contentUrl\":\"https:\/\/umaine.edu\/watershedresearch\/wp-content\/uploads\/sites\/554\/2020\/09\/SDRT_diagram-e1599860134363.png\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/umaine.edu\/watershedresearch\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Research\",\"item\":\"https:\/\/umaine.edu\/watershedresearch\/research\/\"},{\"@type\":\"ListItem\",\"position\":3,\"name\":\"Vulnerability Analysis\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/umaine.edu\/watershedresearch\/#website\",\"url\":\"https:\/\/umaine.edu\/watershedresearch\/\",\"name\":\"Watershed Process and Estuary Sustainability Research Group\",\"description\":\"Web presence of the WPES research group at the University of Maine\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/umaine.edu\/watershedresearch\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Vulnerability Analysis - Watershed Process and Estuary Sustainability Research Group","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/","og_locale":"en_US","og_type":"article","og_title":"Vulnerability Analysis - Watershed Process and Estuary Sustainability Research Group","og_description":"In order to effectively fulfill the mission of managing the health of Maine\u2019s shellfisheries, the Maine Department of Marine Resources requires adequate data and support tools to make scientifically-informed decisions about temporary closures in response to harmful coastal bacteria pollution events. To facilitate data-driven management decisions surrounding closure and re-opening of shellfish flats, School of […]","og_url":"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/","og_site_name":"Watershed Process and Estuary Sustainability Research Group","article_modified_time":"2025-07-10T14:30:10+00:00","og_image":[{"width":284,"height":265,"url":"https:\/\/umaine.edu\/watershedresearch\/wp-content\/uploads\/sites\/554\/2020\/09\/SDRT_diagram-e1599860134363.png","type":"image\/png"}],"twitter_card":"summary_large_image","twitter_misc":{"Est. reading time":"9 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/","url":"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/","name":"Vulnerability Analysis - Watershed Process and Estuary Sustainability Research Group","isPartOf":{"@id":"https:\/\/umaine.edu\/watershedresearch\/#website"},"primaryImageOfPage":{"@id":"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/#primaryimage"},"image":{"@id":"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/#primaryimage"},"thumbnailUrl":"https:\/\/umaine.edu\/watershedresearch\/wp-content\/uploads\/sites\/554\/2020\/09\/SDRT_diagram-e1599860134363.png","datePublished":"2020-07-27T22:42:25+00:00","dateModified":"2025-07-10T14:30:10+00:00","breadcrumb":{"@id":"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/#primaryimage","url":"https:\/\/umaine.edu\/watershedresearch\/wp-content\/uploads\/sites\/554\/2020\/09\/SDRT_diagram-e1599860134363.png","contentUrl":"https:\/\/umaine.edu\/watershedresearch\/wp-content\/uploads\/sites\/554\/2020\/09\/SDRT_diagram-e1599860134363.png"},{"@type":"BreadcrumbList","@id":"https:\/\/umaine.edu\/watershedresearch\/research\/vulnerability-analysis\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/umaine.edu\/watershedresearch\/"},{"@type":"ListItem","position":2,"name":"Research","item":"https:\/\/umaine.edu\/watershedresearch\/research\/"},{"@type":"ListItem","position":3,"name":"Vulnerability Analysis"}]},{"@type":"WebSite","@id":"https:\/\/umaine.edu\/watershedresearch\/#website","url":"https:\/\/umaine.edu\/watershedresearch\/","name":"Watershed Process and Estuary Sustainability Research Group","description":"Web presence of the WPES research group at the University of Maine","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/umaine.edu\/watershedresearch\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"}]}},"taxonomy_info":[],"featured_image_src_large":false,"author_info":{"display_name":"bvandam","author_link":"https:\/\/umaine.edu\/watershedresearch\/author\/bvandam\/"},"comment_info":0,"_links":{"self":[{"href":"https:\/\/umaine.edu\/watershedresearch\/wp-json\/wp\/v2\/pages\/266","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/umaine.edu\/watershedresearch\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/umaine.edu\/watershedresearch\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/umaine.edu\/watershedresearch\/wp-json\/wp\/v2\/users\/1767"}],"replies":[{"embeddable":true,"href":"https:\/\/umaine.edu\/watershedresearch\/wp-json\/wp\/v2\/comments?post=266"}],"version-history":[{"count":50,"href":"https:\/\/umaine.edu\/watershedresearch\/wp-json\/wp\/v2\/pages\/266\/revisions"}],"predecessor-version":[{"id":2274,"href":"https:\/\/umaine.edu\/watershedresearch\/wp-json\/wp\/v2\/pages\/266\/revisions\/2274"}],"up":[{"embeddable":true,"href":"https:\/\/umaine.edu\/watershedresearch\/wp-json\/wp\/v2\/pages\/1862"}],"wp:attachment":[{"href":"https:\/\/umaine.edu\/watershedresearch\/wp-json\/wp\/v2\/media?parent=266"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}