Evolution of the Earth Observation Market

Artist's rendition of a satellite - 3dsculptor/123RF Stock Photo

Artist’s rendition of a satellite – 3dsculptor/123RF Stock Photo

The changing Earth Observation (EO) market has been a topic of office conversation this week at Pixalytics. We’re currently in the final stage of developing our own product portal, and it was interesting to see that some of our thoughts were echoed by reports from last week’s World Satellite Business Week event in Paris.

Unsurprisingly, speakers at the event agreed that the EO sector has huge growth potential. This is something we regularly see highlighted in various emails and press releases. For example, in the last few weeks we’ve had:

At a few thousand dollars for access to each report, we’ve said before that one of the products we should develop is an annual report on the EO market!

As we’ve been working towards our portal, one of issues we’ve identified is how difficult some portals are to navigate, particularly if you are not an EO expert. This was also recognised at the Paris event, with an acknowledgement that EO companies need to understand what customers want and then provide a user friendly experience to deliver those needs.

As reported by Tereza Pultarova in Space News, there was also discussion on the need to move away from simply selling data, and instead provide answers to the practical questions about the planet that businesses and consumers have. It is only through this transformation that new sectors and markets for EO will open which will be the key for the aforementioned future growth. The Paris event also highlighted some of the key trends that will be the backbone of this transformation:

  • Providing as close as possible to near real time data.
  • Increased data analytics, particularly through machine learning and artificial intelligence platforms to analyse data and highlight anomalies and changes faster.
  • Bringing satellite data together with social media information to rapidly enable context to be added to images.
  • Vertical integration within the industry within satellite firms acquiring with data processing and analytics companies; for example, Digital Globe acquired The Radiant Group earlier this year.
  • Processing data onboard satellites, so users download the information they want, rather than reams of data.

There was a really interesting analogy with the navigation industry given by Wade Larson, president and CEO of Urthecast. He said “Navigation became kind of embedded infrastructure in a much larger industry called location-based services. We think that this is happening with geoanalytics.”

This is the direction of travel for the industry, and some players are moving faster than others. Last week Airbus confirmed their four satellite very high-resolution-imaging constellation, Pléiades Neo, is on schedule for launch in 2020. This will have 30 cm spatial resolution and will utilise the Space Data Highway, also known as the European Data Relay System (EDRS), to stream the images into an online platform. The ERDS uses lasers to transfer up to 40 terabytes a day at a speed of up to 1.8 Gbits per second, meaning users will have access to data in near real time.

This evolution of the EO market needs to be recognised by every company in the industry from the Airbus down to the small company’s trying to launch their own product portal. If you don’t move with the changing market, you won’t get any of the market.

China’s Geo-Information Survey

Yuqiao Reservoir, east of Beijing, China from Landsat 8 acquired March 2017. Data courtesy of NASA/USGS,.

The first national geo-information study of China was released last week at a State Council Information Office press briefing.

The study, also referred to as the national census of geographic conditions, was originally announced in March 2013. Over the last three years 50,000 professionals have been involved in collecting a variety of data about China and it’s reported that they have achieved a 92% coverage of the country, generating around 770 terabytes of data in the process.

Data has been collected on natural resources, such as land features, vegetation, water and deserts; together with urban resources such as transport infrastructure, towns and neighbourhoods. This information was gathered, and verified, through remote sensing satellites, drones, aerial photography, 3D laser scanning and in-situ data. It’s reported that the accuracy is 99.7% with a 1 m resolution.

China is one of the largest countries in the world by land mass, at approximately 9.6 m square kilometres. Therefore, simply completing such a study with the accuracy and resolution reported is highly impressive.

It may take years to fully appreciate the variety, size and usefulness of this new dataset. However, a number of interesting high level statistics have already been released by the Chinese Ministry of Land and Resources including:

  • 23.2% of China’s land is above 3,500 m altitude, and 43.4% is below 1,000 m altitude.
  • 7.57 million sq km of the country has vegetation cover, with 21.1% being cultivated lands and the remainder grasslands and forests.
  • 1.3 million sq km of land is desert and bare lands, whilst rivers cover 6.55 million sq km.
  • 153,000 sq km of land has buildings on it.
  • 116,500 sq km of railway track and there is 2 million sq km of roads.

According to Kurex Mexsut, deputy head of the National Administration of Surveying, Mapping and Geoinformation, the Chinese Government will be looking to establish a data sharing mechanism and information services platform for this dataset, together with a variety of data products. It is hoped that public departments and companies will be able to use this to help improve the delivery of public services.

Although not from the survey, the image at the top is of the Yuqiao Reservoir, situated just to the east of Beijing. It has a surface area of 119 sq km, with an average depth of 14 metres.

Not only is this a comprehensive geo-information dataset for a single country, but there is also huge potential for further information to be derived from this dataset. We’ll be watching with interest to see how the data is used and the impact it has.

Supporting Uganda’s Farmers

Map of Uganda showing vegetation productivity. Underlying data is the MODIS 2014 NPP Product, MOD17 – Zhoa et al. (2005).

Uganda is a landlocked country of just over 240,000 square kilometres. Agriculture is a key element of the country’s economy and was responsible for 23% of gross domestic product in 2011 and almost half the country’s exports the following year. According to the Food & Agriculture Organisation of the United Nations, 80% of the population relies on farming for its livelihood.

It has an equatorial climate, with regional variations, although recent recurrent dry spells have impacted on crop and livestock productivity. Pixalytics is delighted to be part of a consortium led by the RHEA Group, working with the Ugandan Ministry of Water and Environment and local NGOs to develop a Drought and Flood Mitigation Service (DFMS) to give practical information to help local communities respond to the effects of climate change.

Using computer models populated with satellite, meteorological, water resources and ground based data an innovative Environment Early Warning Platform will be developed to provide Ugandan farmers, via local NGO organisations, with forecasts throughout the growing seasons to enable them to take actions to maximise their crop yield.

Pixalytics, along with fellow consortium member, Environment Systems, are responsible for the Earth Observation data in the project. We’ll be looking at variety of optical and radar data to provide information about flood and drought conditions alongside crops and their growing conditions.

The project should benefit local communities by:

  • Improving the ability to forecast and mitigate droughts and floods on a local actionable scale.
  • Allowing NGOs to target resources saving time, money and lives.
  • Allowing farmers to improve their lives and better protect their livestock and crops.

Alongside ourselves, and RHEA Group, our consortium includes Environment Systems, Databasix, AA International, AgriTechTalk International, HR Wallingford, UK Met Office, Mercy Corps, and Oxford Policy Management. We will also work with international partners, including the Uganda Government Ministries, Kakira Sugar Company, and the NGO Green Dreams/iCOW. The first of a number of visits to Uganda took place last week, where we had the opportunity to make lots of local contacts and meet some of those whom we hope to benefit from this work.

This work is part of the UK Space Agency’s International Partnership Programme and ours is one of 21 projects chosen to provide solutions to local issues in counties across Africa, Asia, Central and South America.

This is a really exciting project to be involved with, and we’re looking forward to providing useful information to local farmers to allow them to take real and meaningful action to enhance the productivity, and protection, of their livestock and crops.

Uncovering Secrets with Remote Sensing

Artist's rendition of a satellite - mechanik/123RF Stock Photo

Artist’s rendition of a satellite – mechanik/123RF Stock Photo

Recent significant discoveries in Cambodia and Jordan have highlighted the potential offered by remote sensing and satellite imagery to help uncover secrets on Earth – a field known as satellite archaeology.

Cambodia
Helicopter mounted Lidar was used to reveal multiple cities beneath the forest floor near the ancient temples of Angkor Wat in Cambodia. Lidar, which stands for Light Detection and Ranging, is an active optical remote sensing technique that uses a laser scanner to map the Earth’s topography by emitting a laser pulse and then receiving the backscattered signal. In Cambodia, a topographic Lidar with a near infrared laser was used by Australian archaeologist Dr Damian Evans to survey beneath the forest vegetation.

The conurbations discovered, surrounding the stone temple Preah Khan Kompong Svay, are believed to be between 900 to 1 400 years old. Analysis of the survey has shown a large number of homes packed together liked terraced houses, together with structures for managing water and geometric patterns formed from earth embankments – which could be gardens.

At 734 square miles, the 2015 survey is also thought to be the most extensive of its type ever undertaken. Dr Evans work is due to be published in the Journal of Archaeological Science.

Jordan
Archaeologists using high resolution satellite imagery, drones surveys and imagery within Google Earth have discovered a huge structure buried in the sand less than a kilometre south of the city of Petra. The two high resolution satellites used were Worldview-1 and Worldview-2, operated by DigitalGlobe. Worldview-1 was launched in September 2007 and has a half-metre panchromatic resolution; Worldview-2, launched two years later, offers similar panchromatic resolution and 1.85m multispectral resolution.

The outline of the structure measures 56m x 49m, and there is a smaller platform contained inside the larger one. Nearby pottery finds suggest the platform is 2 150 years old, and it is thought that it had a ceremonial purpose. The research undertaken by Sarah Parcak and Christopher Tuttle was published in the May 2016 edition of the Bulletin of the American Schools of Oriental Research.

Benefits of Remote Sensing & Satellites
Angkor Wat and Petra are both World Heritage sites, and the benefits of using remote sensing and satellite technology to undertake archaeological investigations are evident in the statement from Christopher Tuttle who noted that they did not intend to excavate their Petra discovery as ‘The moment you uncover something, it starts to disintegrate.’

Satellite technology allows investigations to take place without disturbing a piece of soil or grain of sand, which is a huge benefit in terms of time, cost and preservation with archaeology. These two discoveries also demonstrate that the world still has secrets to reveal. As Sarah Parcak herself said in 2013, “We’ve only discovered a fraction of one percent of archaeological sites all over the world.”

Who knows what remote sensing and satellite imagery will uncover in the future?

Identifying Urban Sprawl in Plymouth

Map showing urban sprawl over last 25 years in the areas surrounding Plymouth

Map showing urban sprawl over last 25 years in the areas surrounding Plymouth

Nowadays you can answer a wide range of environmental questions yourself using only open source software and free remote sensing satellite data. You do not need to be a researcher and by acquiring a few skills you can the analysis of complex problems at your fingertips. It is amazing.

I’ve been based at Pixalytics in Plymouth, over the last few months, on an ERAMUS+ placement and decided to use Plymouth to look at one of the most problematic environmental issues for planners: Urban Sprawl. It is well known phenomenon within cities, but it can’t be easily seen from ground level – you need to look at it from space.

The pressure of continued population growth, the need for more living space, commercial and economic developments, means that central urban areas tend to expand into low-density, monofunctional and usually car-dependent communities with a high negative ecological impact on fauna and flora associated with massive loss in natural habitats and agricultural areas. This change in how land is used around cities is known urban sprawl.

As a city Plymouth suffered a lot of destruction in World War Two, and there was a lot of building within the city in the 1950s and 1960s. Therefore, I decided to see if Plymouth has suffered from urban sprawl over the last twenty-five years, using open source software and data. The two questions I want to answer are:

  1. Is Plymouth affected by urban sprawl? and
  2. If it is, what are Plymouth’s urban sprawl trends?

1) Is Plymouth affected by urban sprawl?
To answer this question I used the QGIS software to analysis Landsat data from both 1990 and 2015, together with OpenStreetMap data for natural areas for a 15 kilometre area starting from Plymouth’s City Centre.

I then performed a Landscape Evolution analysis, as described in Chapter 9 of the Practical Handbook of Remote Sensing, written by Samantha and Andrew Lavender from Pixalytics. Firstly, I overlaid natural areas onto the map of Plymouth, then added the built up areas from 2015 shown in red and finally added the 1990 built-up areas in grey.

Detailed map showing the key urban sprawl around Plymouth over last 25 years

Detailed map showing the key urban sprawl around Plymouth over last 25 years

The map, which has an accuracy of 80 – 85%, shows you, no major urban development occurred in the city of Plymouth and its surroundings in the last 25 years – this is of course about to change the development of the new town of Sherford on the outskirts of the city.

However, as you can see in the zoomed in version of the map on the right, there is a noticeable urban development visible in the north west of the city and a second in Saltash in Cornwall on the east of the map. The built up area in the 15km area around Plymouth increased by around 15% over the 25 year period. The next question is what are the trends of this sprawl.

2) What are Plymouth urban sprawl trends?
A large amount of research tries to categorize urban sprawl in various types:

  • Compact growth which infill existing urban developments, also known as smart growth, and mainly occurs in planning permitted areas
  • Linear development along main roads
  • Isolated developments into agricultural or wildlife areas in proximity with major roads.

These last two have a bad reputation and are often associated with negative impacts on environment.

Various driving forces are behind these growth types, creating different patterns for cities worldwide. For example, rapid economic development under a liberal planning policy drives population growth in a city which then is expands and incorporates villages located in near or remote proximity over time. This is fragmented approach, and results in a strong land loss.

But this is not the case for Plymouth which in the last 25 years showed a stable development in the extend permitted by planning policies with a predominant infill and compact expansion, a smart growth approach that other cities could take as an example.

These conclusions can be taken following only a few simple steps- taking advantage of free open source software and free data, without extensive experience or training.
This is a proven example of how you can make your own maps at home without investing too much time and money.

This is the end my internship with Pixalytics, and it has been one of my best experiences.

Blog written by Catalin Cimpianu, ERASMUS+ Placement at Pixalytics.

Why Satellite Agri-Tech Applications Will Grow In 2016?

Pixalytics-show preview image2016 is likely to be the year of agri-tech for remote sensing. Its potential has been highlighted for some time, but last year its call was loud and clear.

Agri-tech is the use of technology to improve agriculture production in terms of yield, efficiency and profitability. With a growing global population the need to become more effective and sustainable food producers is obvious, and technology can assist in terms of robotics, biotechnology, navigation, communication, etc. However, it’s opportunities offered by remote sensing that’s most exciting to us – of course, we’re probably biased!

Remote sensing has a wide range of applications for agriculture that range from mapping the underlying soil and crop plus the monitoring of invasive species through to defining seed density optimisation, irrigation management, harvest weather forecasting, yield estimation and long term land change / land use modelling. Essentially, we can offer support from planting to plating!

Despite this potential, uptake within the agricultural sector has been low. A survey of farmers by London Economics / the Satellite Applications Catapult last summer identified barriers that included cost, small-scale justification, reliable mobile / internet signal, lack of software to view data, lack of knowledge and the lack of proven benefits.

So with all of these issues, why are we saying agri-tech will grow in 2016? There are three good reasons:

Benefits Examples – Case studies with concrete examples of the usage of remote sensing are being published. For example, NASA and Applied Geosolutions, worked together using Landsat 8 and MODIS data to examine temperature, greenness, leaf moisture and surface water. This allowed them to develop rice crop management plans, particularly surrounding irrigation, improving both harvest forecasts and actual yields.

Copernicus Sentinel – I know we’ve said this before, but it’s worth saying again, this is a game changer. Both Sentinel-1 and Sentinel-2 data have signals that can be related to vegetation phenology, i.e. how plants change over time. As this data is free, it should allow companies to offer farmers products and services that are not cost prohibitive. Also, as the follow-on missions are launched then the frequency of data coverage will increase – particularly important for optical sensors where clouds can get in the way. Pixalytics has a Sentinel-2 vegetation product in test, which has already been applied to Landsat and very high resolution data, so it’s an area we’re looking to develop further – the image shows a Landsat-8 image processed over land using a Normalised Difference Vegetation Index (NDVI) based algorithm.

Other Data – In June the Department for Environment, Food and Rural Affairs will be making over 8,000 data sets freely available that should cover information such as soil and crop types for fields all over the country. It will provide a wealth of information for farmers to understand what crops they should be growing in which fields to maximise their yields. In addition, the UK’s National Biodiversity Network offers air quality and river level readings.

Taken together these elements offer new opportunities for SME’s to get involved and develop products that will offer real benefits to farmers, both large and small, and will overcome the barriers to them utilising agri-tech. For the right company, with the right idea and right implementation then 2016 will be a high yield year!

If you are interesting in agri-tech and would like to talk to us about what can be done, and what we could offer then please get in touch.

Lidar: From space to your garage and pocket

Lidar data overlaid on an aerial photo for Pinellas Point, Tampa Bay, USA. Data courtesy of the NASA Experimental Airborne Advanced Research Lidar (EAARL), http://gulfsci.usgs.gov/tampabay/data/1_lidar/index.html

Lidar data overlaid on an aerial photo for Pinellas Point, Tampa Bay, USA. Data courtesy of the NASA Experimental Airborne Advanced Research Lidar (EAARL), http://gulfsci.usgs.gov/tampabay/data/1_lidar/index.html

Lidar isn’t a word most people use regularly, but recent developments in the field might see a future where is becomes part of everyday life.

Lidar, an acronym for LIght Detection And Ranging, was first developing in the 1960’s and is primarily a technique for measuring distance; however, other applications include atmospheric Lidar which measures clouds, particles and gases such as ozone. The system comprises of a laser, a scanner and GPS position receiving, and it works by emitting a laser pulse towards a target, and measuring the time it takes for the pulse to return.

There are two main types of Lidar used within remote sensing for measuring distance, topographic and bathymetric; topographic Lidar uses a near infrared laser to map land, while bathymetric Lidar uses water-penetrating green light to measure the seafloor. The image at the top of the blog is a bathymetric Lidar overlaying an aerial photograph Pinellas Point, Tampa Bay in the USA, showing depths below sea level in metres. Airborne terrestrial Lidar applications have also been expanded to include measuring forest structures and tree canopies mapping; whilst there’s ground based terrestrial laser scanners for mapping structures such as buildings.

As a user getting freely accessible airborne Lidar data isn’t easy, but there are some places that offer datasets including:

Spaceborne terrestrial Lidar has been limited, as it has to overcome a number of challenges:

  • It’s an active remote sensing technique, which means it requires a lot more power to run, than passive systems and for satellites this means more cost.
  • It’s an optical system that like all optical systems is affected by cloud cover and poor visibility, although interestingly it works more effectively at night, as the processing doesn’t need to account for the sun’s reflection.
  • Lidar performance decreases with inverse square of the distance between the target and the system.
  • Lidar collects individual points, rather than an image, and images are created by combining lots of individual points. Whilst multiple overflies are possible quickly in a plane, with a satellite orbiting the Earth you’re effectively collecting lines of points over a number of days, which takes time.

The only satellite that studied the Earth’s surface using Lidar is NASA’s Ice, Cloud and Land Elevation Satellite – Geoscience Laser Altimeter system (IceSAT-GLAS); launched in 2003, it was decommissioned in 2010. It measured ice sheet elevations and changes, together with cloud and aerosol height profiles, land elevation and vegetation cover, and sea ice thickness; and you find its data products here. IceSAT-GLAS 2 is scheduled for launch in 2017. The Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO), part of the satellite A-Train, is a joint NASA and CNES mission launched in 2006. Originally designed as an atmospheric focused Lidar, it has since developed marine applications that led to the SABOR campaign we discussed in previous blog.

Beyond remote sensing, Lidar may become part of every household in the future, if recent proof-of-concepts come to fruition. The Google self-drive car uses a Lidar as part of its navigation system to generate a 3D maps of the surrounding environment. In addition, research recently published in Optics Express, by Dr. Ali Hajimiri of California Institute of Technology has described the potential of a tiny Lidar device capable of turning mobile phones into 3D scanning devices. Using a nanophotonic coherent imager, the proof-of-concept device has put together a 3-D image of the front of a U.S. penny from half a meter away, with 15-μm depth resolution and 50-μm lateral resolution.

Lidar has many remote sensing and surveying applications, however, in the future we all could have lasers in our garage and pockets.

The Small and Mighty Proba Missions

This week the European Space Agency announced the latest mission in the Project for OnBoard Automony (PROBA) mini-satellite programme. Proba-3 is planned to launch in four years; and will be a pair of satellites flying in close formation, 150m apart, with the front satellite creating an artificial eclipse of the sun allowing its companion views of the solar corona; normally only visible momentarily during solar eclipses.

Tamar estuary captured in October 2005, data courtesy of ESA.

Tamar estuary captured in October 2005, data courtesy of ESA.

The Proba missions are part of ESA’s In-orbit Technology Demonstration Programme, which focuses on testing, and using, innovative technologies in space. Despite Proba-3’s nomenclature, it will be the fourth mission in the Proba programme. The first, Proba-1, was launched on the 22nd October 2001 on a planned two year Earth observation (EO) mission; however despite the planned lifecycle, thirteen years later it is still flying and sending back EO data. It’s in a sun synchronous orbit with a seven-day repeat cycle and carries eight instruments. The main one is the Compact High Resolution Imaging Spectrometer (CHRIS), developed in the UK by the Space Group of Sira Technology Ltd that was later acquired by Surrey Satellite Technology Limited. CHRIS is a hyperspectral sensor that acquires a set of up to five images of a target, with different modes allowing the collection of up to 62 spectral wavebands.

Plymouth, where Pixalytics is based, and our lead consultant, Dr Samantha Lavender, have a long history with Proba-1. Rame Head point, along the coast from Plymouth, is one of the test sites for the CHRIS instrument and she’s been doing research using the data it provides for over a decade. Over Plymouth Mode 2 is used, which focuses on mapping the water at a spatial resolution of 17m; this mode was proposed by Sam back in the early days of CHRIS-Proba. The image at the top of the page, captured in October 2005, shows the Tamar estuary in the UK that separates the counties of Devon and Cornwall; for this image CHRIS was pointed further North due to planned fieldwork activities. At the bottom of the image is the thick line of the Tamar Road Bridge and below it, the thinner Brunel railway bridge. Plymouth is to the right of the bridge, and to the left is the Cornish town of Saltash.

Proba-V image of the Nile Delta in Egypt, courtesy of the Belgian PROBA-V / ESA Earth Watch programmes

Proba-V image of the Nile Delta in Egypt, courtesy of the Belgian PROBA-V / ESA Earth Watch programmes

Proba-2 was launched in 2009, carrying two solar observation experiments, two space weather experiments and seventeen other technology demonstrations. ESA returned to EO for the third mission, Proba-V, launched on the 7 May 2013; the change in nomenclature is because the V stands for vegetation sensor. It is a redesign of the ‘Vegetation’ imaging instrument carried on the French Spot satellites; it has a 350m ground resolution with a 2250km swath, and collects data in the blue, red, near-infrared and mid-infrared wavebands. It provides worldwide coverage every two days, and through its four spectral bands it can distinguish between different types of land cover. The image on the right is from Proba-V, showing the Nile delta on 2nd May 2014.

Despite their small stature all the Proba satellites are showing their resilience by remaining operational, and they’re playing a vital role in allowing innovative new technologies to be tested in space.

Science won’t be rushed!

Ocean currents derived from GOCE data. Image courtsey of ESA/CNES/CLS

Ocean currents derived from GOCE data
Image courtsey of ESA/CNES/CLS

Last week scientists presented the most accurate model of ocean currents created to date using data from the GOCE satellite; twelve months after the satellite burnt up on re-entry to the Earth’s atmosphere.

Immediacy is the standard bearer for the twenty-first century; social media allows everyone to tell the world the instant they’ve done anything, emails must be read as soon as they come in and there’s the frustration you feel when a web page takes a few seconds to load. Science doesn’t easily fit this world. The scientific method takes time; it’s about developing a hypothesis, experimenting to test that hypothesis, analysing the data and comparing it the original idea, and then often tweaking the hypothesis and repeating the cycle. Instant gratification is rarely found in this methodology, as shown by the GOCE work.

GOCE, the Gravity field and steady-state Ocean Explorer, was launched in 2009. It was five metre long and was known as the ‘Ferrari of space’ due to its sleek design and the fact it was assembled in Italy; and was the lowest flying scientific satellite. Its main instrument was an Electrostatic Gravity Gradiometer which measured gravity gradients in all directions. It flew for four years running out of fuel in October 2013, and returned to earth a few weeks later.

GOCE’s instrumentation mapped minute changes in Earth’s gravity which were used to construct a ‘geoid’ – a hypothetical global ocean at rest – with an accuracy of 1-2cm at a 100km resolution. This geoid model was subtracted from the mean sea-surface height measured over a 20-year period by other satellites, including Envisat, to create a map of the ocean surface showing areas of water higher, and lower, than average. The map was then used to calculate ocean currents and speeds, and was validated through in situ measurements. The resulting model of ocean currents was presented at the 5th International GOCE User Workshop last week.

GOCE, like all satellites, collected a huge amount of data, which takes time to analyse. A similar example is likely to be the Philae probe which landed on comet 67P/Churyumov-Gerasimenko last month. During the sixty hours it was operational after landing; it sent streams of measurements back to Earth. Even without future communication with the probe, this data will take years to fully analyse, study and interpret. Who knows what scientific discoveries may be made from this?

You can’t rush science. It takes time, and effort, to get accurate and reliable scientific results. But those results have the potential to change our understanding of the world and help create innovations to improve the future of our planet and our lives. Now that’s worth waiting for isn’t it?

Smashing the Earth Observation Data Silos

Artist's rendition of a satellite - paulfleet/123RF Stock

Artist’s rendition of a satellite – paulfleet/123RF Stock

Earth observation (EO) is an all-encompassing term for monitoring our world, however as soon as you start examining the topography of the field in detail you’ll find all sorts of mountains, valleys and oceans. An illustration of the different stands can be seen if you consider the subject areas such as hydrography, geology, surveying and remote sensing, or think about areas of interest like the land and the marine specialists, and finally think about sensors specialists for LIDAR, optical or hyperspectral imaging. Historically a lot of these groupings have tended to work in relative isolation with a limited amount of interaction between them, which has created a lot of EO data, and knowledge, silos. However as satellite technology has developed, the quantity of EO data available has increased exponentially; for example, Landsat is currently collecting fourteen times as many images each day than it was in the 1980’s. Whilst many datasets have been collected, few have been brought together. This is due to both computing power required to manage large datasets and the difficulties of cross-calibrating sensors with different errors and uncertainties. Cloud computing has broken through most of the data processing obstacles, giving the potential for many more people to get involved in data manipulation, modelling and visualisation. The next challenge is to smash open these data silos, and provide access to historical archives, and new collections, to both the scientific community, and anyone else who is interested. Joining together the different strands of data and knowledge will promote innovation and help us significantly develop our understanding of the planet. Individual space agencies are working on this through making new data freely available and by analysing their own historical archives and then reprocessing them to improve consistency. Some examples include:

Progress is being made, but there are still limitations as often this only represents the bringing together of data from a single mission; a product set or thematic group. There is a need to be bolder and to amalgamate much wider datasets. Last week, Taiwan demonstrated how this could be achieved by presenting their petascale database for assessing climatic conditions, which has brought together data from the atmosphere, hydrology, ocean currents, tectonics and space. The Earth Science Observation Knowledge base holds ten and half million records and gives scientists near real time access to data. EO has a vast array of valuable data and is collecting more every day. We’re starting to smash the data silos, but we need to do more to achieve the next step change in understanding how our world works.