EnxAero:
You can see the terrain going up 2000 m before the threshold but the chart ends before the peak is reached. 2000 m before the threshold, the terrain is still below the airport elevation. I guess it is not common to have a relevant feature 3km before the threshold...but on the other hand, I guess it is not common to have a RA of 500' at DH either.
The 2000m limit is required by ICAO rules.
Another way to look at it: from ICAO's perspective for the purposes of using RA to determine DH, anything over 2 km from the threshold should not be considered relevant.
There could be rising terrain under the approach 3 km away, 5 km away, right below the FAF, etc., but as long as they are below the protection surface, they should not be considered for evaluating the DH RA.
roulette:
It's still rather strange that the ILS Cat I DAs for Cat BCD are higher than the LOC Only MDA!
It is not very common, but does happen. For another example look at the runway 16R approaches into Reno, Nevada (KRNO). There the
ILS DA at 2031 ft AGL is a whopping 460 ft (!) higher than the
LOC MDA.
As mentioned earlier this can happen when there are obstacles penetrating the ILS protection surface.
For non-precision approaches, the approach design can use step-down fixes (including the MDA floor) to basically "clear" any obstacles.
For precision approaches, the design must either raise the DA or raise the approach angle beyond 3.00 degrees.
Since protection must be extended below the glide-slope and below the DA in case of missed approaches, the raised DA could well be above the MDA.
(For LECO, a hint here is the LOC CDA angle being 3.10 degrees while the ILS angle is 3.00 degrees). In this case, the choice of raising the DA vs. raising the ILS angle is unknown to me.