Imagery and the search for MH370

It’s one thing collecting imagery, altogether another to utilize it.  Digital Globe’s WorldView satellite can collect as much as 750,000 square kilometres a day and revisit the same place on the earth’s surface every 1.7 days.  There are several satellites with similar characteristics and so the theoretical re-collects are very frequent.  So even with the vast search area for MH370, the imagery can be collected.  All this really means is that we have a picture of the vast haystack.

Weather is an obvious challenge.  The only sensors with the resolution needed to identify debris are optical sensors.  Cloud or fog renders them useless.  Sadly, the weather in the Southern Ocean rarely cooperates – the pale areas in this application hint at real-time cloud cover:,-33.83,1289.

But collecting imagery is never where the real bottleneck is – collecting the picture of the field of haystacks is easy, finding the needle isn’t.  This is where much of the misunderstanding of imagery comes from…  ‘seeing’ an image (as in the contextual imagery we see in Google Maps or other consumer mapping applications) is straightforward but ‘searching’ an image isn’t.

Finding objects in a featureless sea might seem straightforward.  The issues are that the sea is vast, is not featureless (‘white caps’ look just like debris) and the debris from an aircraft is small.   There are sophisticated algorithms available to find ships in imagery; there are no algorithms available to find more randomly shaped debris.  The only tool that can reliably identify debris is the human eyeball and the brain.  That then leads to the problem of how many eyeballs are available.

Within the imagery intelligence community, the ‘normal’ problem is lessened by ‘cueing’ a sensor to a specific area of interest, often identified by other intelligence assets.  Even then, there are very specific signatures that are being sought;  automatic processing is realistic.  The speculation in the media that the intelligence community is all-seeing is simply wrong.  In this case, what you’re seeing in the media from commercial sources isn’t greatly different to what’s behind the security curtain.  This is not a ‘normal’ intelligence problem.

​Indeed, the commercial sector is ahead of the game in one crucial area:  crowdsourcing.  Digital Globe’s Tomnod platform ( allows thousands of eyeballs to search imagery and tag items of interest.   This provides an outlet for frustrated citizens to contribute to the search for MH370.   Even so, Tomnod can only support the assessment of a limited number of images – it still requires cuing in.  Frustratingly, Tomnod hasn’t yet got imagery of the Southern search area… perhaps illustrating that the scale of the problem may be too great even for Tomnod.

Any of these approaches introduce significant delay between the acquisition of the imagery and the finding of objects of interest.   We’ve already seen the 3-5 day lags in the ‘official’ images that have been released from US, Chinese and French sources.  Sadly, in the Southern Ocean, Current and Wind move objects a long way in that time.  Surface winds can be seen at,-33.83,1289 and ocean currents at,-33.83,1289.  Those currents in the search area are particularly problematic – they’re all over the place!

Dropping tracking buoys helps to identify future drift but that presupposes that aircraft have sighted the debris spotted in the satellite imagery.  So far, that hasn’t happened.

A major challenge in tracking debris is that the interaction between ocean current and surface wind is based on how the debris is floating.  It it’s almost entirely submerged, the ocean current effect dominates.  If there is a substantial amount above the surface, the winds will have a major effect.  Sadly, the resolution of the imagery will never allow the debris to be understood at this level.  Spotting debris from aircraft will improve that understanding… in terms of plotting the future track of the debris.  But from this you’ll appreciate that knowing where debris spotted from satellite five days earlier is now… is almost impossible.

Sadly, even assuming that debris from MH370 is found in the Southern Ocean, a major challenge remains… tracking back to where it originated so that the search for the black boxes can commence.   As you can see from the ocean current map above, the chaotic nature of the ocean in the search area suggests a long search ahead.

We hope this blog helps to explain some of the challenges and frustrations faced by the searchers and through clarifying some of the complications brings a level of comfort through these testing times.