Open Imagery: Why freely available satellite data is so important for science

Landsat satellite conceptualisation*

An article released in Nature by Gabriel Popkin (2018), referenced the value of freely available access to the United States' Landsat missions archive - which is the longest continuous data series that we have available in the remote sensing community - and raised the concern that the current U.S. administration were considering whether to reintroduce charges to access this imagery. This is not something new, as the Obama administration assessed whether there would be a benefit to charging for the satellite sensor imagery during their first term and they deemed this to be inappropriate when considering the vast amount of scientific output that is generated due to the accessibility of these data.

However, the world has benefited from access to these data for a decade now and anyone can easily access almost 50 years worth of satellite imagery. This latest review is was announced just after the tenth anniversary of free access in which the USGS stated that 'as of March 31, 2018, more than 75 million Landsat scenes have been downloaded from the USGS-managed archive!' (USGS 2018).

There are a huge number of stakeholders that benefit from the ability to freely download satellite sensor imagery, including government departments from across the world (particularly those dealing with planning, agricultural and emergency response), academics (in a number of disciplines and at various stages of the career ladder), NGOs (the practice of using remotely sensed imagery is becoming more common within a human rights context), new technology and geospatial businesses which may not have started trading if the charges to imagery were still present (Fernholz 2018) and news organisations among many others. One particular group who hugely benefit from the ability to download this imagery for free are students.

For many people University is the first experience of remote sensing and satellite imagery analysis for students outside of using Google Earth. The ability to explore an archive of more than 40 years of satellite imagery from the Landsat mission is a great tool to introduce theories of spectral, spatial and temporal analysis that can be applied to a variety of research questions. Before the opening of the archive many students relied on the use of imagery that had been purchased through the research grants their lecturer's had secured but the ability to download free medium-resolution imagery allowed for more variety in the materials that lecturer's had access to for their classes and has meant that undergraduate students have the ability to conduct their own dissertation using remotely sensed techniques without requiring funding for the research.

Should the Landsat series once again have a cost associated this would vastly limit the amount of research that could be conducted over large time-scales which is counterproductive particularly from an environmental perspective as we look to temporal changes in the past to help us model the possible effects to ecology and the wider environment from large-scale events such as climate change. Research has immensely benefited from the removal of charges for the satellite imagery with a huge influx of new research projects as the feasibility of carrying out remote sensing studies became more realistic as it was no longer required that the funding grant should be spent on imagery, but rather this could be used for other supporting elements of the research enabling the development of more detailed research projects and an advancement of the field of remote sensing and photogrammetry.

Reintroducing charges to the procurement of Landsat imagery would be devastating to the remote sensing community, and although there is a commitment from the European Space Agency (ESA) to allow for free and open access to their data from the Sentinel satellite series which is very useful and will be an important source of imagery continuing in to the future, the risks to the sector by reintroducing charges to Landsat still remain. Sentinel data is compatible with the Landsat sensors, but has major improvements in terms of revisit times and pixel size (Drusch et al. 2012). The Sentinels are the first high-resolution satellite sensors to provide free data to the masses for research purposes. However, despite all the benefits that are available with the Sentinel programme, they have not been collecting data long enough to allow for significant temporal analyses, and it will be a very long time until the levels of data available from these sensors matches those collected by Landsat, and it will never be able to observe the decades that were available before their launch.

It would be an absolute disaster for the continued development and progress of remote sensing sciences. Significant advances have been made in the community since the release of Landsat data free of charge in 2008, and without this, many of the strides that have been made would cease to exist. (In 2001 the USGS sold 25,000 Landsat images which increased to a distribution of more than 1 million in 2008 when the data became freely available to the public - Borowitz 2018a). A recent study found that data from more than half of unclassified Earth-observing satellites is restricted in some manner, rather than being openly shared (Borowitz 2018b), and should the Landsat archive was to be added to this list the consequences for scientific discovery would be dire. As there is a continued push for transparent and open access data repositories, research publications and open access programming, charging for a publicly funded satellite programme that easily pays for itself due to the subsequent research output would be a huge step backwards. Landsat data are used around the world, it was a pioneering venture in remote sensing and is perhaps the most famous and most well loved. Hopefully the current U.S. administration will realise the benefits that the Landsat series provides to people around the world in the same way others have, but until then scientists around the world much continue to voice their concerns about free access to data and demonstrate why this is so important.

The wheels appear to be set in motion however, as the USGS recently sent out a survey to Landsat users requesting people's opinions regarding the type of data they wish to see collected on the proposed Landsat 9 mission. Throughout completion of the survey I had a foreboding sense that this was really to scout out how much people would be willing to pay for access to data that has been free for the past decade to the great benefit of science. Unfortunately I was not proved wrong as there were questions regarding the Landsat user community's willingness to pay for the data, as was mentioned earlier in the year when the review was announced (Pixalytics 2018). Although I hope the U.S. federal budgets will be amended to prevent the need for charges to be applied to imagery downloads, I fear for both the access to free imagery from Landsat and the continuation of the longest running EO program in history, particularly as the European Space Agency has committed to its free and open access imagery policy (Trovatello 2017), particularly with the Copernicus program, which includes their Sentinel-2 satellites.

The USGS has set up a page which answers some of the questions on the current status of the review including a link to the National Geospatial Advisory Committee website which will post periodic updates on the process and also provides an email for people to share their experiences of Landsat and the ability to access these data freely which may help with future Landsat data policies. You can find all of this information here.

*Downloaded from Google (Images) labeled for noncommercial reuse with modification, with the Search term 'landsat satellite', original image available via NASA Goddard Space Flight Center's photostream on flickr. Available: https://www.flickr.com/photos/gsfc/7630269434 [Accessed 06.07.2018].


References 

Borowitz, M. (2018a). Half of Earth's satellites restrict use of climate data. The Conversation. Available http://theconversation.com/half-of-earths-satellites-restrict-use-of-climate-data-93257 [Accessed 06.07.2018].

Borowitz, M. (2018b). Open Space: The Global Effort for Open Access to Environmental Data. MIT Press, Cambridge MA.

Drusch, M., Del Bello, U., Carlier, S., Colin, O., Fernandez, V., Gascon, F., Hoersch, B., Isola, C., Laberinti, P., Martimort, P., Meygret, A., Spoto, F., Sy, O., Marchese, F. and Bargellini, P. (2012). Sentinel-2: ESA's Optical High-Resolution Mission for GMES Operational Services. Remote Sensing of Environment, 120, 25-36. 

Fernholz, T. (2018). The US government wants to start charging for the best free satellite data on earth. Quartz. Available https://qz.com/1278239/the-us-government-wants-to-start-charging-for-landsat-the-best-free-satellite-data-on-earth/ [Accessed 06.07.2018].

Pixalytics. (2018). Will you have to pay for Landsat Data? Available https://www.pixalytics.com/pay-for-landsat-data/ [Accessed 20.09.2018].

Popkin, G. (2018). US government reviews data fees. Nature, 556, 417-418.

Trovatello, M. (2017). ESA affirms open access policy for images, videos and data. European Space Agency. Available www.esa.int/About_Us/Digital_Agenda/ESA_affirms_Open_Access_policy_for_images_videos_and_data [Accessed 20.09.2018]. 

USGS. (2018). April 20, 2018 - 10th Anniversary of Imagery for Everyone. Landsat Missions, United States Geological Survey. Available https://landsat.usgs.gov/april-20-2018-10th-anniversary-imagery-everyone [Accessed 06.07.2018].

Comments