tillergirl
Well-Known Member
I think crowd source data should be banned. It is difficult because it can add some very good information. But the trouble is the absence of metadata. Any comment in that context must have full metadata and even then I think that is risky when a user's understanding of navigation is based on an iphone.I'd actually regard that as dangerous. Recontouring the bathymetric data is fraught with potential for error, as frequently seen with Navionics crowd sourced data, where missing data is interpreted as licence to interpolate, rather than "no-one goes there so it isn't safe"!. The contours provided in UKHO data are based on far more data than is present on the chart, and are reliable and drawn so that any error will be on the safe side of indicating shallower water than it actually is. But recontouring on the basis of the selected soundings on the chart and the contours is not a safe or straightforward matter; there are many possible algorithms and they all give different answers! The only one where you get an estimate of the accuracy of the interpolation is Kriging, which is extremely computing intensive, even now with modern processors, as it has to examine all the data to create a semi-variogram.
I should have added to my previous that one manufacturer several years back did add a drying height in a channel but did not draw the contours for a month (it might be more but I need to go back to my records to check). So during that month if you missed that single drying height .........

