Short communication
A dynamic graphical interface for visualizing projected, measured, and reconstructed surface water elevations on the earth's largest lakes

https://doi.org/10.1016/j.envsoft.2013.07.003Get rights and content

Abstract

There is a growing need within the international water research and water resources management community, and the general public, for easy access to time-series of projected, measured, and reconstructed marine and freshwater coastal surface water elevations. There is also a need for effectively communicating variability among different surface water elevation data sets, as well as the intrinsic uncertainties in surface water elevation forecasts. Here, we introduce an interactive web-based interface, the Great Lakes Water Level Dashboard (GLWLD), designed to address this need for the North American Laurentian Great Lakes, the largest assemblage of unfrozen fresh surface water bodies on planet Earth, and one with a coastline of over 16,000 km (roughly 10,000 miles). The GLWLD is a Flash-based tool that can simultaneously display time-series of measured monthly and annual water level data and seasonal forecasts for each of the Great Lakes, reconstructed lake levels from paleoclimate research, and decadal lake level projections under alternative climate scenarios. By employing a suite of novel data transfer, processing, and visualization tools, the GLWLD allows users to seamlessly transition not only between alternate displays of Great Lakes water levels over different temporal scales, but between different data sets and forecasts as well. Furthermore, the unique GLWLD interface can help users understand the extent to which decisions regarding the use of the lakes depend on an appreciation of uncertainty and variability within, and between, different sources of Great Lakes water level information.

Introduction

The North American Laurentian Great Lakes collectively constitute the largest surface area and the second largest volume of any unfrozen surface freshwater resource on planet Earth, and are directly and critically linked to the human, environmental, and economic health of central North America (Buttle et al., 2004, Millerd, 2005, Field et al., 2007). As a massive and dynamic inland coastal system, the Great Lakes respond to a combination of intrinsic and extrinsic forces ranging from climate change impacts to anthropogenic controls (Brown et al., 2011). This range of drivers propagates into a variety of large-scale physical and ecological features. Of these, changes in the surface water elevations of the lakes themselves are arguably one of the most important from a regional water resources planning perspective.

Existing web-based outlets for Great Lakes water level information cover a range of spatial and temporal scales. Dynamic displays of hourly and daily-scale data and forecasts, for example, are available from the National Oceanic and Atmospheric Administration (NOAA) Great Lakes Coastal Forecasting System (GLCFS) and the Great Lakes Observing System (Schwab and Bedford, 1994, Read et al., 2010). Similarly, static images of both official and experimental monthly-scale water level forecasts are distributed by, respectively, the Detroit District of the United States Army Corps of Engineers (USACE) and the NOAA Great Lakes Environmental Research Laboratory (GLERL). These existing federal institution frameworks underscore the importance of effectively communicating Great Lakes water level information to the general public, policy makers, and regulatory authorities. There is, therefore, a clear need for a single interactive portal through which users can access and analyze time-series data and forecasts of Great Lakes water levels from multiple sources across various temporal scales.

We find, furthermore, that most Great Lakes water level forecasts (and forecasts for a variety of other meteorological, economic, and climate information) are presented to the general public without a basis for assessing their accuracy and the extent to which that accuracy (or model skill) varies over annual and sub-annual time scales (Spiegelhalter et al., 2011). The USACE, for example, routinely publishes operational Great Lakes seasonal water level forecasts in the Monthly Bulletin of Lake Levels for the Great Lakes along with summary statistics from the observed water level record. This bulletin provides the general public with a simple, internationally-coordinated forecast and is a readily-available source of accurate and up-to-date information on Great Lakes monthly-average water levels. However, the bulletin (as with most static images) does not give users the option of overlaying archived forecasts and observations so that they get a sense of the forecast model's success rate. This type of retrospective model assessment, when easily accessible, allows researchers to prioritize investments in model improvements while potentially helping the general public understand relationships between risk, forecast-based decision making, and forecasting skill (Murphy, 1993). To address this need, scientists from NOAA-GLERL and the University of Michigan's Cooperative Institute for Limnology and Ecosystems Research (CILER) developed the Great Lakes Water Level Dashboard (GLWLD). This web-based graphical user interface employs state-of-the-art software technology to display and compare Great Lakes monthly and annual lake-wide average water level time-series data, and forecasts, from multiple sources.

Section snippets

Overview of GLWLD data and forecasts

The GLWLD is a versatile tool designed to communicate monthly, annual, and decadal Great Lakes water level data and forecasts to a broad user community ranging from recreational boaters, marina owners, and hydropower facility managers, to representatives from the Great Lakes shipping industry and coastal infrastructure design teams. As described in the following subsections, the GLWLD organizes this information into four categories.

System design

The GLWLD is designed to build on the success of other graphical user interface-based data analysis tools (Hyman et al., 1996, Jeong et al., 2006) while complying with guidance proposed by the environmental data visualization community (Yi et al., 2007, Kelleher and Wagener, 2011). A design attribute of many innovative web-based data access tools is a display that incrementally unveils additional “levels” of information to the user (Spiegelhalter et al., 2011). Implementing this design feature

Representative application

The GLWLD was designed to improve how Great Lakes monthly, seasonal, and decadal water level data and forecasts are communicated to a broad audience, and to move beyond the conventional protocol of disseminating individual “pre-generated” static graphics through multiple web sites. One example of how the GLWLD can be used to achieve this goal is through display of long-term (i.e. multi-decadal) Great Lakes water level projections.

Long-term Great Lakes water level projections have gained

Conclusions and future work

The GLWLD is a novel and freely-available web-based tool that utilizes state-of-the art data visualization and processing schemes to communicate important information about water levels of the North American Laurentian Great Lakes. Many of the data sets in the GLWLD are derived from readily-available web sites and static images that, while informative, do not facilitate a direct comparison between different data sets and forecasts of the same variable across a variety of time scales. The GLWLD

Acknowledgments

Funding for this project was provided by NOAA, the Great Lakes Restoration Initiative (administered by the United States Environmental Protection Agency), and the International Joint Commission (IJC) International Upper Great Lakes Study. The authors thank NOAA's Climate.gov dashboard team, including Viviane Silva, David Herring, and Mark Phillips, for consultation on the initial development of the GLWLD, as well as Margaret Lansing, Glenn Muhr, and Cathy Darnell (all from NOAA-GLERL) for

References (40)

  • S.J. Baedke et al.

    A 4,700-year record of lake level and isostasy for Lake Michigan

    Journal of Great Lakes Research

    (2000)
  • C. Brown et al.

    A decision-analytic approach to managing climate risks: application to the upper Great Lakes

    Journal of the American Water Resources Association

    (2011)
  • J. Bunch

    Mission of US Lake Survey

    Journal of the Surveying and Mapping Division

    (1970)
  • J. Buttle et al.

    Economic impacts of climate change on the Canadian Great Lakes hydro-electric power producers: a supply analysis

    Canadian Water Resources Journal

    (2004)
  • A.H. Clites et al.

    Midlakes, a Coordinated Hydrologic Response Model for the Middle Great Lakes

    (1998)
  • T. Croley

    Long-term heat storage in the Great Lakes

    Water Resources Research

    (1992)
  • T. Croley

    Mixing probabilistic meteorology outlooks in operational hydrology

    Journal of Hydrologic Engineering

    (1997)
  • T. Croley

    Weighted-climate parametric hydrologic forecasting

    Journal of Hydrologic Engineering

    (2003)
  • J. Curry et al.

    Climate science and the uncertainty monster

    Bulletin of the American Meteorological Society

    (2011)
  • C. Field et al.

    North America

  • Cited by (28)

    • Development of a preliminary vegetation-based indicator of ecosystem health for coastal wetlands of the Laurentian Great Lakes

      2020, Ecological Indicators
      Citation Excerpt :

      Data collection for the Coastal Wetland Monitoring Program (CWMP) began in the summer of 2011 and has been conducted every summer since, therefore providing an optimal opportunity to identify vegetation-based metrics throughout a sequence of low to high water level years. In 2011 and 2012, the Great Lakes were experiencing a prolonged period of historically low water levels, particularly in lakes Huron-Michigan and Superior (Gronewold et al., 2013). By 2014, the lakes had transitioned to a period of higher water levels, 10–20 cm above long-term water levels for lakes Huron-Michigan and Superior, continuing at least through 2019.

    • Real-time environmental sensor data: An application to water quality using web services

      2016, Environmental Modelling and Software
      Citation Excerpt :

      Data systems employed across the environmental domains may be broadly classified into two groups: 1) systems used for the storage, retrieval and visualization of data, and 2) data systems designed explicitly for real-time operations. While the former do not explicitly treat real-time data, they do provide powerful mechanisms by which to standardize data retrieval and storage (Christodoulou et al., 2010, Horsburgh et al., 2009, Argent et al., 2009, Gronewold et al., 2013, Goodall et al., 2008). Some of these platforms conform to a set of community standards (e.g. WaterML, DelftFEWS, etc., see (Taylor et al., 2010)) that reduce operational overhead and enable the seamless use of standard-compliant tools for scalable storage, management and visualization of data.

    • An expandable web-based platform for visually analyzing basin-scale hydro-climate time series data

      2016, Environmental Modelling and Software
      Citation Excerpt :

      As an example, instead of focusing on sub-basins within the Great Lakes (see top of Fig. 2), the GLD-TSCOM could work with hydrological inputs, outputs, and storage within subregions of California, defined by the United States Geological Survey (USGS), currently experiencing extreme drought conditions (Swain et al., 2014). The GLD and the GLD-TSCOM were preceded by the Great Lakes Water Level Dashboard, or GLWLD (Gronewold et al., 2013) and the Great Lakes Hydro-Climate Dashboard, or GLHCD (Clites et al., 2014). Prior to their release, many previously available data services advertised as Great Lakes data or containing Great Lakes data were exclusive to the United States, leaving out Canadian data, which are invaluable for Great Lakes basin-wide data and model output.

    View all citing articles on Scopus
    View full text