The MT Blog

Updating Italy

11 October 2017
Simon, G4TJC


For a very long time we have been well aware of the poor accuracy of the SOTA summits lists for the Italian association. Horizontal accuracy was pretty hopeless, sometimes kilometres off, with many summits apparently in valleys even. Also only a small fraction of the eligible summits were included. This meant that most of the easier summits were not included in the programme, closing the activity off to those less able to tackle the higher summits.

Serious work to correct this situation started over two years ago. Techniques were developed first on the IS0 (Sardinia) association before moving on to the I association.

After retiring 352 summits and adding 3466 new references the I association now had 3666 active in its listing.

Some of the new references replace an earlier reference for the same summit. The team did try to identify old and new data to pair, so that the earlier reference would be kept. However, it became apparent that this was too hard a task. The older data were, in many cases, simply too ambiguous, and efforts were better spent more productively. This means that for a few summits which were already activated under an old reference it will be possible again to claim the first activation using the new reference!

So, what was our methodology for evaluating the new listings?


First some principles...

  • SOTA in Italy is to be treated the same as any other association
  • The criterion for listing is P ≥ 150 m
  • Because the summit density is high, there is no derogation to P100 status
  • There are to be no exclusions on the basis of some elevation cut-off
  • There are to be no exclusions on the basis of restricted access or anything else
  • The best data available to the MT will be used


The actual techniques used were broadly as outlined in my earlier blog postings. This means that the team first identified candidate summits using computer analysis of a digital elevation model.

Our normal data source for this step is SRTM, collected by NASA using the Space Shuttle. However, in the case of Italy we actually had better data. These are from TINItaly, an analysis by Istituto Nazionale di Geofisica e Vulcanologia:

A TIN is actually a triangular irregular network. Here the technique has been used to combine various sources into one consistent model. I feel I should point out that to get access you have to request it, and I'm sure the people there would prefer not to receive lots of new requests to process all of a sudden. You can be sure that the team has processed the data coherently, providing a list of candidates for the next step.


10-m contours from TINItaly (somewhere pretty steep!)

This takes us to map checking. Each candidate is found on the best available map, checking summit and col. Even when we think the P is somewhat short of the requirement we check anyway.

Italy's national mapping agency is Istituto Geografico Militare. They release topographic maps online. This is what the team has used.

Spot heights are used where available. Otherwise we have to use the contours. This will be the highest contour at the summit and the lowest contour just above the col. Using these values results in a so-called clean prominence. It means that you can be confident that any summit listed really does conform to the minimum P criterion. It does mean some marginal summits don't make the list, whereas in fact an accurate survey would show that they would pass. But we would rather not list a handful of good summits than list "bad" ones, only to have to de-list them later. We are open to reconsidering marginal, excluded summits when supported by credible data.


Time Wasting (yours and ours)

... here are some of the things we would consider too dubious.

  • Google Earth - Yes, we use Google Earth as a useful tool. When we are short of better data we might even use it to support (or not) summit entries, comparing, for example, with SRTM and whatever maps we can find. However, in this case we have TINItaly. Google Earth adds nothing to our knowledge.
  • Open Street Map (in its various renderings) likewise tells us nothing new about the elevations. SRTM data are already folded in to TINItaly with appropriate weighting.
  • Recreational GPS units or phone GPS - elevation measurements in particular have poor repeatability and should be given little weight compared with other sources. Maybe this will change if accuracy improves substantially (Galileo perhaps?)
  • Other data of unknown provenance

But we would like...

  • Better IGM mapping than is freely available
  • Proper surveys
By that I mean surveys in the style of the Relative Hills of Britain group, whose surveys form the basis of our listings for the UK and Ireland. If you want to set up for this you will need to find a few €k for survey-grade instruments! Also you will need to dedicate several hours to measuring each summit / col.


We are aware that some of the summit names may not be the best. We are open to input. This should be sent via the Italian AM, once this appointment has been confirmed.