Supporting Soil Fertility From Space

Sentinel-2 pseudo-true colour composite from 2016 with a Kompsat-3 Normalized Difference Vegetation Index (NDVI) product from 2015 inset. Sentinel data courtesy of ESA/Copernicus.

Last Tuesday I was at the academic launch event for the Tru-Nject project at Cranfield University. Despite the event’s title, it was in fact an end of project meeting. Pixalytics has been involved in the project since July 2015, when we agreed to source and process high resolution satellite Earth Observation (EO) imagery for them.

The Tru-Nject project is funded via Innovate UK. It’s official title is ‘Tru-Nject: Proximal soil sensing based variable rate application of subsurface fertiliser injection in vegetable/ combinable crops’. The focus is on modelling soil fertility within fields, to enable fertiliser to be applied in varying amounts using point-source injection technology which reduces the nitrogen loss to the atmosphere when compared with spreading fertiliser on the soil surface.

To do this the project created soil fertility maps from a combination of EO products, physical sampling and proximal soil sensing – where approximately 15 000 georeferenced hyperspectral spectra are collected using an instrument connected to a tractor. These fertility maps are then interpreted by an agronomist, who decides on the relative application of fertiliser.

Initial results have shown that applying increased fertiliser to areas of low fertility improves overall yield when compared to applying an equal amount of fertiliser everywhere, or applying more fertiliser to high yield areas.

Pixalytics involvement in the work focussed on acquiring and processing, historical, and new, sub 5 metre optical satellite imagery for two fields, near Hull and York. We have primarily acquired data from the Kompsat satellites operated by the Korea Aerospace Research Institute (KARI), supplemented with WorldView data from DigitalGlobe. Once we’d acquired the imagery, we processed it to:

  • remove the effects of the atmosphere, termed atmospheric correction, and then
  • converted them to maps of vegetation greenness

The new imagery needed to coincide with a particular stage of crop growth, which meant the satellite data acquisition period was narrow. This led to a pleasant surprise for Dave George, Tru-Nject Project Manager, who said, “I never believed I’d get to tell a satellite what to do.’ To ensure that we collected data on specific days we did task the Kompsat satellites each year.

Whilst we were quite successful with the tasking the combination of this being the UK, and the fact that the fields were relatively small, meant that some of the images were partly affected by cloud. Where this occurred we gap-filled with Copernicus Sentinel-2 data, it has coarser spatial resolution (15m), but more regular acquisitions.

In addition, we also needed to undertake vicarious adjustment to ensure that we produced consistent products over time whilst the data came from different sensors with different specifications. As we cannot go to the satellite to measure its calibration, vicarious adjustment is a technique which uses ground measurements and algorithms to not only cross-calibrate the data, but also adjusts for errors in the atmospheric correction.

An example of the work is at the top, which shows a Sentinel-2 pseudo-true colour composite from 2016 with a Kompsat-3 Normalized Difference Vegetation Index (NDVI) product from 2015 inset. The greener the NDVI product the more green the vegetation is, although the two datasets were collected in different years so the planting within the field varies.

We’ve really enjoyed working with Stockbridge Technology Centre Ltd (STC), Manterra Ltd, and Cranfield University, who were the partners in the project. Up until last week all the work was done via telephone and email, and so it was great to finally meet them in-person, hear about the successful project and discuss ideas for the future.

UK Space Conference Getting Ready For Take Off

Next week we’ll be in Manchester at the 2017 UK Space Conference.

The UK Space Conference is held every two years, and attracted over 1,000 delegates and over 100 exhibitors when held in Liverpool in 2015. It is a key event that brings together the UK Space Community and this year is taking place over three days, 30th May to the 1st June.

We are exhibiting on stand C7, near the centre of the hall, where you’ll be able to come and talk to us about our products and services including:

  • Atmospheric correction
  • Consultancy services
  • Education & training
  • Flood mapping
  • Ocean colour
  • Spatial analyses & data management
  • Terrestrial vegetation
  • Turbidity mapping

We’re also delighted to announce that our Flood Mapping work is one of the products highlighted in the Innovation Zone, which is sponsored by Innovate UK. It is a low cost floodwater mapping product based on Sentinel-1 radar data, which provides easy to understand flood information and maps through an online portal without the need for specialist knowledge. We have partnered with Harris Geospatial Solutions to provide a fully automated solution.

We’ll also have copies of our book for sale, ‘Practical Handbook of Remote Sensing’. This takes complete novices through the process of finding, downloading, processing, visualising and applying remote sensing satellite data using their own PC, open-source software and a standard internet connection.

The 2017 UK Space Conference itself begins on the Tuesday morning with ‘Space 101’, which is a series of workshops covering some of the key issues related to working in the space sector. The conference then kicks off at lunchtime on the Tuesday with an opening plenary on the latest developments in the UK space sector.

There is a networking event in the Exhibition Hall between 6pm and 9pm on Tuesday evening, and we’ll be on our stand all evening.

Wednesday is brimming over with workshops, presentations, plenary and poster sessions, culminating in the Gala Dinner and Sir Arthur Clarke Awards. Finally, Thursday has another busy day of workshops and plenary sessions, before the Conference closes in the afternoon.

We’re really excited about being in Manchester next week, and looking forward to meeting old and new friends.

We hope that any of you who at the Conference will come up and say hello! We’d love to meet you!

Goodbye to EO-1

Hyperspectral data of fields in South America classified using Principle Components Analysis. Data acquired by Hyperion. Image courtesy of NASA.

In contrast to our previous blog, this week’s is a celebration of the Earth Observing-1 (EO-1) satellite whose death will soon be upon us.

EO-1 was launched on the 21st November 2000 from Vandenberg Air Force Base, California. It has a polar sun-synchronous orbit at a height of 705 km, following the same orbital track as Landsat-7, but lagging one minute behind. It was put into this orbit to allow for a comparison with Landsat 7 images in addition to the evaluation of EO-1’s instruments.

It was the first in NASA’s New Millennium Program Earth Observing series, which had the aim of developing and testing advanced technology and land imaging instruments, particularly related to spatial, spectral and temporal characteristics not previously available.

EO-1 carries three main instruments:

  • Hyperion is an imaging spectrometer which collects data in 220 visible and infrared bands at 30 m spatial resolution with a 7.5 km x 100 km swath. Hyperion has offered a range of benefits to applications such as mining, geology, forestry, agriculture, and environmental management.
  • Advanced Land Imaging (ALI) is a multispectral imager capturing 9 bands at 30 m resolution, plus a panchromatic band at 10 m, with a swath width of 37 km. It has the same seven spectral bands as Landsat 7, although it collects data via a different method. ALI uses a pushbroom technique where the sensor acts like a broom head and collects data along a strip as if a broom was being pushed along the ground. Whereas Landsat operates a whiskbroom approach which involves several linear detectors (i.e., broom heads) perpendicular (at a right angle) to the direction of data collection. These detectors are stationary in the sensor and a mirror underneath sweeps the pixels from left to right reflecting the energy from the Earth into the detectors to collect the data.
  • Atmospheric Corrector (LAC) instrument allows the correction of imagery for atmospheric variability, primarily water vapour, by measuring the actual rate of atmospheric absorption, rather than using estimates.

The original EO-1 mission was only due to be in orbit only one year, but with a sixteen year lifetime it has surpassed all expectations. The extension of the one year mission was driven by the Earth observation user community who were very keen to continue with the data collection, and an agreement was reached with NASA to continue.

Psuedo-true colour hyperspectral data of fields in South America. Data acquired by Hyperion. Image courtesy of NASA.

All the data collect by both Hyperion and ALI is freely available through the USGS Centre for Earth Resources Observation and Science (EROS). At Pixalytics we’ve used Hyperion data for understanding the capabilities of hyperspectral data. The two images shown in the blog are a subset of a scene acquired over fields in South America, with image to the right is a pseudo-true colour composite stretched to show the in-field variability.

Whereas the image at the top is the hyperspectral data classified using a statistical procedure, called Principle Components Analysis (PCA), which extracts patterns from within the dataset. The first three derived uncorrelated variables, termed principle components, are shown as a colour composite.

Sadly, satellites cannot go on forever, and EO-1 is in its final few weeks of life. It stopped accepting data acquisition requests on the 6th January 2017, and will stop providing data by the end of February.

It has been a great satellite, and will be sadly missed.

Sentinel-3 Sets Sail

Artist's view of Sentinel-3. Image courtesy of ESA–Pierre Carril.

Artist’s view of Sentinel-3. Image courtesy of ESA–Pierre Carril.

At 17.57 GMT yesterday (16th February 2016) Sentinel-3 set sail from the Plesetsk Space Centre in Russia, heading for its 814 km sun-synchronous low Earth orbit. Like all the other Sentinel launches, we were at home watching the live feed!

This is the third Sentinel launch of the European Commission’s Copernicus Programme, following Sentinel-1 and 2. Sentinel-3, like its predecessors, will be part of a twin satellite constellation with Sentinel-3B’s launch expected to be in 2017.

Sentinel-3 carries four scientific instruments:

  • Sea and Land Surface Temperature Radiometer (SLSTR) will measure temperatures of both the sea and land, to an accuracy of better than 0.3 K. This instrument has 9 spectral bands with a spatial resolution of 500 m for visible/near-infrared wavelengths and 1 km for the thermal wavelengths; and has swath widths of 1420 km at nadir and 750 km looking backwards. It’s worth noting that two thermal infrared spectral wavebands are optimised for fire detection, providing the fire radiative power measurement.
  • Ocean and Land Colour Instrument (OLCI) has 21 spectral bands (400–1020 nm) focussed on ocean colour and vegetation measurements. All bands have a spatial resolution of 300 m with a swath width of 1270 km.
  • Synthetic Aperture Radar Altimeter (SRAL) which has dual frequency Ku and C bands. It offers 300 m spatial resolution after SAR processing, and is based on the instruments from the CryoSat and Jason missions. This will be first satellite altimeter to provide 100% coverage of the Earth’s surfaces in SAR mode.
  • Microwave Radiometer (MWR) dual frequency at 23.8 & 36.5 GHz, it is used to derive atmospheric column water vapour measurements for correcting the SRAL instrument.

The scientific instruments are supported by four positioning/navigation instruments to ensure the satellite maintains its precise orbit.

Sentinel-3 will mainly be focussing on ocean measurements and will include the measurement of sea-surface height (similar to the recently launched Jason-3); however it will also measure sea surface temperature, ocean colour, surface wind speed, sea ice thickness and ice sheets. Whereas over land the satellite will provide indices of vegetation, measuring the height of rivers and lakes and help monitor wildfires.

Sentinel-3 is a very exciting satellite for us, as the data and products it will produce are very much within the wheelhouse of the services that Pixalytics offers. Sam’s background is in ocean colour, she’s world renown for atmospheric correction research and we offer a variety of agritech services including vegetation indices. You can probably now see why we’re so excited!

The satellite is currently in its commissioning phases where ESA tests the data produced by the sensors. This is undertaken in conjunction with a group of users, and Pixalytics is one of them! This phase is expected to last five months, after which the satellite will be transferred to Eumetsat and the data should be released.

Like all the data from the Copernicus programme, it will be offered free of charge to users. This will challenge organisations, like us, to see what innovative services we can offer with this new data stream. Exciting times ahead!

Current Work in Remote Sensing and Photogrammetry

Last week the annual Remote Sensing and Photogrammetry Society (RSPSoc) conference was held in Aberystwyth. Now I’ve stepped down as RSPSoc Chairman I could relax and enjoy this year’s event as a delegate.

Arriving on Wednesday morning, the first session I attended was organised by the Technology and Operational Procedures Special Interest Group (TOPSIG), which was focused on Operational Earth observation. There were a great range of presentations, and I particularly enjoyed the user insights by Andy Wells on how customers are really using imagery. Recent developments in on-the-fly importing, georeferencing and autocorrelation means bringing data together from different sources isn’t a time consuming chore. Users can therefore spend more time analysing data, extracting information and adding value to their organisations or research. In addition, as highlighted by other presentations, open software repositories continue to grow and now include complex algorithms that were once only available to specialists. Finally, Steve Keyworth reminded us that what we do should be seen as a component of the solution rather than the specification; the ultimate aim should be on solving the customer’s problem, which in the current climate is often financially motivated.

Landsat 7 image showing features in the Baltic, data courtesy of ESA

Landsat 7 image showing features in the Baltic, data courtesy of ESA

On Thursday I co-chaired the Water and Marine Environments session alongside Professor Heiko Balzter, on behalf of the Marine Optics Special Interest Group (SIG). My presentation was focused on the European Space Agency (ESA) Landsat archive that’s been acquired via the ESA ground stations. This data is being reprocessed to create a consistent high resolution visible and infrared image dataset combining the three primary sensors used by the series of Landsat satellites; MSS (Multi-spectral Scanner), TM (Thematic Mapper), and ETM+ (Enhanced Thematic Mapper Plus). Although historical Landsat missions are not ideally suited to observing the ocean, due to a low signal-to-noise ratio, features can be clearly seen and the new processing setup means images are being processed over the open ocean.

Mark Danson’s keynote lecture on Friday morning described the application of terrestrial laser scanners to understanding forest structure. He showcased his post PhD research which has led to the development of the Salford Advanced Laser Canopy Analyser, a dual-wavelength full-waveform laser scanner. The presentation also showed the importance of fieldwork in understanding what remote techniques are actually sensing, and in this case included a team of people cutting down example trees and counting every leaf!

Mark also made me feel less guilty that I am still working on a component of my PhD – atmospheric correction. In research your own learning curve, and the scientific process, mean you gain new insights as you understand more, often explaining why answers are not as simple as you might have assumed. It’s one of the reasons why I love doing research.

Overall, I had a great time at RSPSoc, catching up and seeing what’s new in the field. My next conference event is Ocean Optics, in the US, at the end of October where I’ll be discussing citizen science in a marine science context.

Home from Hawaii

I got back to a ‘cold’ UK on Saturday afternoon after spending last week at Ocean Sciences 2014.  It was a fantastic conference with over 5,600 attendees.  My scientific highlights were:

The Surface Ocean Layer Atmosphere Study (SOLAS) session on Monday where speakers presented research on the sea surface microlayer (the top 1 mm of the ocean); this layer is important so we can understand the transfer of compounds, such as carbon dioxide, and particles from the ocean to the atmosphere and vice versa that are critical to our interpretation of the climate.

On Tuesday afternoon it was the Optics and Light in the Particle-Laden Coastal Ocean session, with presentations focused on understanding the acoustic and optical signatures of particles, including their shape, from multi-angular measurements and Lidar (laser) profiling of a phytoplankton bloom.

My key session was obviously Optical Remote Sensing of Freshwater, Estuarine and Coastal Environments on Wednesday. I gave a presentation on Multi-Sensor Ocean Colour Atmospheric Correction for Time-Series Data.  Atmospheric correction is the removal of the atmosphere’s signal from data so only the water-leaving radiance signal is left; it allows data to be compared between days irrespective of the weather conditions of that day – so an image taken on a hazy day will look like it was taken on a clear day.

HICO™ Data, image of Hong Kong from the Oregon State University HICO Sample Image Gallery, provided by the Naval Research Laboratory

HICO™ Data, image of Hong Kong from the Oregon State University HICO Sample Image Gallery, provided by the Naval Research Laboratory

Other interesting talks from this session included Tiit Kutser’s presentation on comparing in-situ measurements with MERIS data for dissolved organic carbon and iron concentrates in Lake Malaren in Sweden, Keping Du’s retrieval algorithm for phycocanian, a pigment within cyanobacteria, within Taithu lake in China, Heidi Dierssen’s optics of seagrass for remote sensing and I also really enjoyed my mentee Guangming Zheng’s presentation on suspended sediment within Chesapeake Bay, off the west coast of America – this took me back to my PhD that focussed on the suspended sediment plume from the River Humber.

Finally, there were great presentations by Curt Davis and Nick Tufillaro on the Hyperspectral Imager for the Coastal Ocean (HICO) mission. It’s an experimental mission that’s designed to sample the coastal ocean; one 50 x 200 km scene per orbit at a spatial resolution of around 90 m. The image on the right shows a HICO example.

On top of these oral sessions, I also spent time in the exhibition, poster sessions and some of the evening events.  My last event on the Thursday evening was about getting involved in the European Commission’s Horizon 2020 Research programme – so if anyone needs an Earth Observation specialist partner for their bid, get in touch!

Settling in at the 2014 Ocean Sciences Meeting

My beach reading spot!

My beach reading spot!

I’m feeling a bit guilty writing this with everyone in the UK still experiencing the wet wintery weather, as I’m in Honolulu, Hawaii at the 2014 Ocean Sciences Meeting. I arrived around midnight last Friday and spent Saturday morning relaxing on the edge of Waikiki beach looking out to sea, watching people go by and reading a book; and I’d recommend Mr Penumbra’s 24-Hour Bookstore by Robin Sloan as a great story if you’re interested technology, cryptography or old books!

Ocean Sciences 2014 opened with a welcome reception on Sunday evening and an interesting key note address by Polynesian explorer Elizabeth Kapu’uwailani Lindsey entitled ‘Bringing the Wisdom of the “Elders” Together with Modern Science for the Future of the Environment’.  During the evening I did try to try track down the three people I’m mentoring at this conference – Elaine, Guangming and Navid. However, as there are over 5,000 attendee’s there were a lot of people to search through!

I finally sat down with my mentees at breakfast on Monday morning. The meeting’s mentoring program is an interesting concept designed to help novice conference attendees get the most value out of their experience.

Yesterday we met up at Navid’s poster to discuss his work on modelling internal waves, at lunch we discussed Elaine’s poster on climate data services and tomorrow Guangming is presenting a talk on the turbidity maximum in Chesapeake Bay in the same session as I’m presenting my ocean colour atmospheric correction work. This is a great element about going to scientific conferences, meeting people to discuss their research and interests. Some are new acquaintances and some I’ve known for a long time. My first time attendance at this meeting (also in Hawaii) was in 2006.

Today I’ve also been to a session on particles in the coastal ocean, and this evening it was Philanthropic Investment in Ocean Research. Tomorrow I’ve got a really busy day as I’m speaking, I’ll give you an update on the second half of the conference next week!

A little addendum I noticed while finishing this post is that Satellite Applications Catapult yesterday published its response to UK Space Innovation and Growth Strategy Space Growth Action Plan. I’ve downloaded it, and it will form part of my next beach reading session!