Page 22 - GIS for Science, Volume 3 Preview

INTEGRATED AND MULTIDISCIPLINARY COASTLINE CLASSIFICATION
The classification approaches described earlier generally focus on a single classification variable, whether it is environmental sensitivity to a disturbance, tectonic setting, erosional versus depositional nature, or dominant hydrodynamic influence. In recognition of the strong physical, ecological, and human diversity associated with coastal environments, there have also been important attempts to classify coastal environments using multiple factors. Cooper and McLaughlin19 presented a review of emerging coastal classification approaches and argued that advances in spatial analytical technologies (GIS and remote sensing) enabled increased focus on multifactor classification at that time. Evaluating and quantifying the classic “holistic” view of coastlines as integrated products of geomorphological and oceanographic processes7 was more easily accomplished using geographic analysis and spatial technologies. These emerging technologies allowed for complex multidisciplinary analysis and mapping of the many physical, ecological, and human aspects of coastal environments. For example, Jelgersma et al.20 classified low-lying deltaic areas using 18 variables, which were grouped into four main categories: Offshore water environment (six marine variables); Coastline properties (seven shoreline morphology variables); Deltaic system properties ( four variables describing land and river conditions); and Human activity (one variable). This study was an early example of a complex delta classification approach aimed at including “everything that matters.”
Globally comprehensive, standardized, integrated, high spatial resolution charac- terizations and maps of coastal systems are still relatively lacking. At the turn of the 21st century, a major World Resources Institute analysis concluded, “Information on the location and extent of coastal features and ecosystems types often provides the basis for subsequent analyses of condition of the ecosystem, relationships between different habitats, and overall trends. Yet, despite this fundamental importance, such information is incomplete and inconsistent at the global level.”4 In response to many similar calls for the detailed mapping of Earth’s ecosystems, the Group on Earth Observations (GEO), a consortium of more than 100 nations and participat- ing organizations seeking to advance the use of earth observation for environmental problem-solving, has commissioned a task (task T1 in the 2020-2022 GEO Ecosys- tems Implementation Plan published at www.earthobservations.org) to produce a standardized, robust, and practical classification and map of the planet’s terrestrial, freshwater, and marine ecosystems. The work we have done to characterize coastal ecosystems, described next, was undertaken in response to that official commission from GEO and complements our earlier related GEO work to characterize global pelagic ecological marine units (EMUs).21
Our approach—coastal segment units (CSUs) and ecological coastal
units (ECUs)
Our team produced a standardized, consistent, high spatial resolution, and globally comprehensive characterization of the coastlines of the world to integrate coastal water properties, coastal land properties, and properties of the coastline itself. The team partitioned the coastlines of the world into 4 million 1-km or shorter segments and attributed those segments with data on 10 variables that describe the basic ecological settings in which the coastline segments occur. The partitioning and attribution have resulted in a new open data resource with which any 1-km coastline segment, anywhere in the world, can be queried, and the values for 10 ecologically meaningful characteristics of that segment returned.
Our primary objective in undertaking this work was to produce globally comprehensive data that describe the ecological settings of coastlines. A 1 km coastline segment may be an appropriate spatial resolution and analytical unit for coastal managers. As such, our team hopes that while the characterization is global, the resulting product may have management utility at place-based scales. Moreover, it appeared that 1 km is a large enough distance to aggregate summaries of other environmental data, yet small enough to show rich variety in those summaries. Integrating the data on features of the land, water, and coastline showed the potential of spatial data integration. The data product is intended to be useful to managers as a extensive inventory of the ecological settings of coastal areas. Every 1-km segment with a unique set of class labels for the 10 attributes is called a coastal segment unit (CSU). The data are also intended to be useful for coastal zone research, including coastal zone assessments related to the Sustainable Development Goals (SDGs) of the UN 2030 Agenda for Sustainable Development. (Readers can find a link to this document and other digital resources relating to the work at GISforScience.com.)
A secondary objective in undertaking this work was to identify groups of similar coastal areas based on their aggregate ecological setting, as determined from a global statistical clustering of all the segments data. The “all-in” clustering was performed with all 10 attributes included and with equal weighting of the attribute variables. The global clustering was exploratory in nature, and the preliminary results distinguished 16 distinct coastline environments. These 16 distinct coastline environments are called ecological coastal units (ECUs). The ECUs represent broad groups of globally similar coastal environments.
10
GIS for Science
Remote coastline, Eucla, Western Australia.