- We now have better observations capabilities at satellite, aerial, UAV, and ground levels
- Cube satellites are enabling small satellites to be accessible by many
- Move to open data also enables a wide range of data to be accessible
- Artificial intelligence and ‘deep learning’ techniques increasingly used for analysis
Transformations in Remote Sensing and Earth Observation
Remote sensing and earth observation have radically transformed in the last decade or so, as a variety of new observation platforms have become available, while scientists and researchers have been inundated with new data.
Remote sensing, that is remote observation from platforms that are distant to the objects being observed, and Earth observation, which is the gathering of information on the physical, biological, and related data about the planet, are two complementary and distinct fields that have similar interests and research commonalities.
Remote sensing has seen new platforms that cover different levels of observation, from high to low altitudes. Levels of observation have included satellite data, aerial data, including aircraft data, unmanned aerial vehicles (UAV) data, and ground-based observation (e.g., geophysical equipment).
Earth observation can include systems that measure the environment and surroundings, including temperature and wind gauges among other monitoring devices.
Recent trends in remote sensing and earth observation include manufacturers increasingly bringing systems together, such as Light Detection and Ranging (Lidar) being integrated with satellite, aerial, and UAV platforms.
Free weekly newsletter
Fill out your e-mail address to receive our newsletter!
By entering your email address you agree to receive our newsletter and agree with our privacy policy.
You may unsubscribe at any time.
As Aliastair Graham, in this podcast, points out, scientists have increasingly seen their needs require multiple levels of observation, driving increased capabilities in data and computing systems that handle such data while also creating remote sensing platforms that can integrate instruments initially created for other platforms.
Similar change includes how cube satellites (CubeSats) have developed in recent years as platforms of 10 x 10 x 10 cm miniature satellites that can be put together more easily and cheaply, while such satellites can also work together in constellations or groups.
Companies such as Capella Space and Planet Labs have delivered these cube satellites where these satellites can be made to work together, even with other groups. In recent years, we have also seen the larger satellite systems such as Sentinel 1 and 2 working more together.
With the near ubiquitous presence of UAVs, we also now have a variety of observations from high and low altitudes available to us.
Open Data and Open Source Initiatives
Although there have been extraordinary developments by private satellite companies, many other key developments have come from governments in Europe, North America, and Asia that have made satellite data either free or very cheap.
Aliastair Graham, in another podcast (podcast), has highlighted how open satellite data initiatives such as the Copernicus dataset, Landsat, and Sentinel satellite data, along with open software that has developed to process remote sensing data, have created a situation where many researchers feel overwhelmed with data. It also means many remote sensing specialists no longer need to purchase data.
Even some private companies have begun releasing their large archives or at least part of their archival data, such as RadSat.
With such data becoming common, remote sensing specialists will find themselves increasingly depending on machine learning and deep learning techniques that can handle large and varied remote sensing datasets.
Raster Vision is an example of an open source initiative that uses deep learning and machine learning techniques to help researchers process remote sensing data. Despite what might be seen as a financial cost to companies and government by releasing data for free, public and private interests have seen benefits by opening data, such improved environmental observation (e.g., monitoring reduced ice and its impacts) or even commercial benefits and forecasting. By releasing their data for free, the user community has greatly expanded and more platforms applying analyses have become available.
Better Search Capabilities
With the great deluge of data becoming present, and much of it for free, new companies have emerged that specialize in finding relevant data for users.
Companies such as Apollo Mapping and SkyWatch have specialized in finding the right type of data for needs companies and others have.
Formats for data capture have also gained traction in the remote sensing and earth observation industries, including enhanced compression wavelet (ECW), NetCDF, and GeoTIFFs, with GeoTIFFS still generally the most common industry standard.
The Open Geospatial Consortium (OGC ) has become increasingly important as web services and data are provided as a key service, with a variety of standards becoming increasingly common (e.g, SOS), including web map service (WMS).
Both front-end and back-end developments in analysis and indexing of data have seen improvements in recent years, such as the creation of the OpenEO initiative that allows open projects based on Python, R, and other software and computer languages to better work with data.
The Future
What we can expect in the future, and in fact what has already begun, is that companies include more web processing services.
With developments in Cloud Computing and services, data processing is becoming a key service focus.
Users mainly want an answer to research or market questions. Web services can provide that answer by processing different remote sensing data, irrespective of the platform, and provide that data result rather than only a type of remote sensing data (e.g,. Landsat or Lidar).
Users are also keeping their data online as Cloud services improve, with only the outputs provided as download.
This is also likely to change the model in how users gain access to data. Rather than requesting types of satellite data, users subscribe to objects (e.g., houses, cars) that they are interested in monitoring and then they get that result based on the object, with the service analyzing different remote sensing data in the background.
Questions, rather than data type, are likely to drive how people access information, which means we also need better questions that are specific to the needs of researchers and industry.
Researchers will also need to work better together to improve the variety and scale of data becoming available. This includes better tools and questions to coordinate activities but also collaboration among specialists in aerial, UAV, ground-based, and satellite remote sensing communities.
Computer scientists and data science skills are becoming critical within remote sensing as the field transitions to an effectively geospatial data science.
Despite these challenges, it is clear that more and varied data will mean we can better observe the Earth in the years to come.
