What do the intertidal maps depict?

The maps were created to identify the non-vegetated areas of Earth's coastline that undergo regular tidal inundation. In some areas, these occur as tidal flats up to 24-km wide, such as the tidal mudflats of western Europe and East Asia. Our analysis included 56 predictor layers of which many were Landsat composite metrics designed to identify individual pixels that undergo frequent wetting and drying.

How accurate are the maps?

Our approach achieved >82% accuracy when compared to independent, globally distributed, validation data. In many areas on earth pixels undergo a similar wetting and drying regime, so it is possible to find areas of aquaculture and coastal development that display similar dynamics. In addition, areas where there were few satellite images available for the analysis, or where the water is highly turbid, may also cause commission error.

Can I use the data for any purpose?

The data and analysis code are made available as open access with appropriate acknowledgement. This work is licensed under a Creative Commons Attribution 4.0 International License. Use of any aspect of this study must include appropriate acknowledgement, which includes citing the paper (see below), and reporting the data source and version.

How do I access and analyse the data?

The data can be analysed directly within Google Earth Engine or downloaded from the outlets listed on the download page.

How do I cite the data?

Please cite the paper: Murray N. J., Phinn S. R., DeWitt M., Ferrari R., Johnston R., Lyons M. B., Clinton N., Thau D. & Fuller R. A. (2019) The global distribution and trajectory of tidal flats. Nature. 565:222-225. http://dx.doi.org/10.1038/s41586-018-0805-8

How do I analyse the data as time-series?

Owing to the variable availability of Landsat images over the study period, each time step in the intertidal change data has a varying extent. Therefore, to analyse the data as a time series it is necessary to take extreme care. You can use the QA layers to develop masks of the minimum extent where the classifier was implemented. Furthermore, as a time-series where each time-step has variable accuracy, we recommend using an appropriate statistical model rather than direct measures of area change. Please read the paper or contact the corresponding author for further information.

Who can I contact to discuss the study?

Please contact the corresponding author of the paper (Nicholas Murray).

Are there any tutorials on how to access and use the intertidal data?

We have developed several Google Earth Engine tutorials to show how to access, analyse and download the data.

  • Download the data using Google Earth Engine code editor (link).
  • Develop QA masks for downloading a spatially consistent time series (link).
  • Contribute training data for use in version 2 using Google Earth Engine code editor (link).
  • Contribute training data for use in version 2 using the free, open source remote sensing app, Remap (link).
  • Contribute training data as comma separate value files (link).

For more information on learning to use Google Earth Engine refer to the Google Earth Engine website.