3 ms·

According to the recovered black box recordings, one of the crew did see the island, but not soon enough. Probably the most shocking part is that there was a l

According to the recovered black box recordings, one of the crew did see the island, but not soon enough. Probably the most shocking part is that there was a lighthouse on this island, which was operational at the time. This is shocking both because it should have been blindingly obvious to the pilot and crew that they were flying toward a lighthouse, and also that an island significant enough to have a lighthouse on it would not be on a digital mapping system. It's on Google maps - https://goo.gl/maps/Xdt7rzUErX42 https://goo.gl/maps/Xdt7rzUErX42, even has photos and information. This is not some insignificant rock How an aircraft mapping system can have such a huge omission is just mind-blowing

theoh 10y ago
To be clear, the deficiency was in a warning system database of 3D elevation geometry (the EGPWS) not a "mapping system". Something that is geographically notable but quite small, vertical and isolated (like Rockall) may very well also be omitted from that version of the EGPWS database. NB Blackrock is a designated navigational waypoint on the standard charts etc. which is why they were flying straight toward it on autopilot.
- roryisok 10y ago
  > Something that is geographically notable but quite small, vertical and isolated (like Rockall) may very well also be omitted from that version of the EGPWS database. Why are "small" and "isolated" good criteria for omitting something from a database of elevation geometry? It's evidently big enough to crash into and cause loss of life.
  - theoh 10y ago
    I'm not suggesting that those are criteria, just possible contributing factors. (And I'm not trying to exonerate the suppliers. We'll see what the final report says.) If the source is some kind of DEM (digital elevation model) data, it will be a raster of elevation values at some limited resolution, so necessarily just samples of elevation. With a relatively small island/rock, the full height could be missed, and possibly then even further reduced by filtering (i.e. smoothing of DEM values, either in the process of measurement or some kind of processing). That likelihood is hinted at by this quote from the manufacturer: "Honeywell’s terrain data is sourced from our supplier [named supplier]. It is a digitized topographic map derived data set. It does not include Black Rock. We have looked at alternate sources, including SRTM and ASTER. While Black Rock is present in these alternate data sets, the actual altitude of Black Rock is considerably higher than what is indicated in these alternate data sets." The data sources they refer to: http://www2.jpl.nasa.gov/srtm/ http://www2.jpl.nasa.gov/srtm/ https://asterweb.jpl.nasa.gov/gdem.asp https://asterweb.jpl.nasa.gov/gdem.asp Again, not coming to any conclusions from this.
Declanomous 10y ago
Lighthouses are built for ships, not helicopters though. You can have a landmass that would be rather significant for a ship but not at all significant for an airplane or helicopter; think a reef or ridge which never rises above the waves.

theoh 10y ago

To be clear, the deficiency was in a warning system database of 3D elevation geometry (the EGPWS) not a "mapping system". Something that is geographically notable but quite small, vertical and isolated (like Rockall) may very well also be omitted from that version of the EGPWS database. NB Blackrock is a designated navigational waypoint on the standard charts etc. which is why they were flying straight toward it on autopilot.

roryisok 10y ago
> Something that is geographically notable but quite small, vertical and isolated (like Rockall) may very well also be omitted from that version of the EGPWS database. Why are "small" and "isolated" good criteria for omitting something from a database of elevation geometry? It's evidently big enough to crash into and cause loss of life.
- theoh 10y ago
  I'm not suggesting that those are criteria, just possible contributing factors. (And I'm not trying to exonerate the suppliers. We'll see what the final report says.) If the source is some kind of DEM (digital elevation model) data, it will be a raster of elevation values at some limited resolution, so necessarily just samples of elevation. With a relatively small island/rock, the full height could be missed, and possibly then even further reduced by filtering (i.e. smoothing of DEM values, either in the process of measurement or some kind of processing). That likelihood is hinted at by this quote from the manufacturer: "Honeywell’s terrain data is sourced from our supplier [named supplier]. It is a digitized topographic map derived data set. It does not include Black Rock. We have looked at alternate sources, including SRTM and ASTER. While Black Rock is present in these alternate data sets, the actual altitude of Black Rock is considerably higher than what is indicated in these alternate data sets." The data sources they refer to: http://www2.jpl.nasa.gov/srtm/ http://www2.jpl.nasa.gov/srtm/ https://asterweb.jpl.nasa.gov/gdem.asp https://asterweb.jpl.nasa.gov/gdem.asp Again, not coming to any conclusions from this.

Declanomous 10y ago

Lighthouses are built for ships, not helicopters though. You can have a landmass that would be rather significant for a ship but not at all significant for an airplane or helicopter; think a reef or ridge which never rises above the waves.