5 replies to this topic

[Clark Geomatics]

Hi All,

I'm in the process of generating a series of maps for mountaineering / backcountry skiers (see attached). I'd really appreciate some feedback on this work in progress - so far I've generated the landuse types from satellite imagery and created contours that delineate glaciers / topography. The points on the map are temporary and are from GPS waypoints captured during hikes / traverses - I use these to create the routing / trails.

The final map will be printed at 33" x 30" - map on one side and supporting information and photos on the back.

The PNGs - CartoTalk_1.png is 332.7k / CartoTalk_2.png is 422.5k - show different views of the map (overview and more detailed) - I can re-post if the resolution is lacking.

Thanks for taking the time to reply.

Cheers,

Jeff

Attached Files

•  CartoTalk_2.png   422.53K   354 downloads
•  CartoTalk_1.png   332.7K   316 downloads

[frax]

looks very nice! I wouldn't mind seeing a higher res image...

[Martin Gamache]

I moved this to the Gallery section.


Jeff,

Can you talk about the data you are using?

The Landcover classes look very nicely done. I like the palette alot. Is this from Landsat? How are you doing the classification? The transition from glacier to non-glacier is very sharp. I would suggest adjusting that through some use of texture or other class in your landcover. The glacier/nonglacier boundary is probably the hardest thing to depicty correctly on this sort of map due to the mixed pixels and all the things going on there like shadows, moraine, scree... It is something I have struggled with alot. Your boundary seems very clean which is misleading....

How will the GPS data be incorporated into the final maps? The straight lines currently linking those points are distracting and probably not useful from a navigational perspective. maybe for skiing they are fine.

This brings me to my last point. Have you considered producing different versions of your map for skiers and hikers. The reason I ask is that we always tend to depict landcover as summer or late fall. The winter/spring landscape that skiers will encounter will be very different and the map may not be as useful...something to think about.

A higher res version so we can look at the type would be nice

BTW You don't say where this is...looks like Garibaldi Park to me...

[Clark Geomatics]

Thanks for the comments Martin - see below for my replies.

Can you talk about the data you are using?  The Landcover classes look very nicely done. I like the palette a lot. Is this from Landsat? How are you doing the classification?

<{POST_SNAPBACK}>

Various data sources were used to generate the map - the elevation model is from Canadian CDED data; the hydrography is partially from Landsat 7 and partially from a provincial dataset; the land classes were derived from a supervised Landsat 7 classification and further tweeked in Photoshop.

The transition from glacier to non-glacier is very sharp. I would suggest adjusting that through some use of texture or other class in your landcover. The glacier/nonglacier boundary is probably the hardest thing to depicty correctly on this sort of map due to the mixed pixels and all the things going on there like shadows, moraine, scree... It is something I have struggled with a lot. Your boundary seems very clean which is misleading....

<{POST_SNAPBACK}>

Right - I know exactly what you mean - it's a tough area to deal with spectrally without touching up in Photoshop. This is understandable when you look at a photograph of the area - the contrast between the bright snow and grey / black rock is drastic. So, in previous areas, I've extracted a 'texture' from the Landsat 7 pan band to merge with the glaciated areas - it works nicely to define the glacier and blend it in a little with the surrounding morraine / scree / etc.

The other reason the glacier / non-glacier border is quite sharp is because of the contour / glacier outline being a little heavier that the regular topo isolines - I can tone down the glacier isolines and emphasize the topo isolines so the don't 'clash' as much.

How will the GPS data be incorporated into the final maps? The straight lines currently linking those points are distracting and probably not useful from a navigational perspective. maybe for skiing they are fine.

<{POST_SNAPBACK}>

You are right - they are only there temporarily as a guide for capturing skiing and hiking routes. The points / lines will be replaced with curved arrows and other navigational symbols. I'll post something when I get that far.

This brings me to my last point. Have you considered producing different versions of your map for skiers and hikers. The reason I ask is that we always tend to depict landcover as summer or late fall. The winter/spring landscape that skiers will encounter will be very different and the map may not be as useful...something to think about.

<{POST_SNAPBACK}>

I've thought about that a lot - I've talked to quite a few hardcore backcountry folks and they seem to think that as long as they know what is glaciated, where the tree-line is located and what are the access roads, they are happy (they are not a fussy bunch - they just want to ski). But I think you have a good point - a Landsat7 captured in the early spring would probably be beneficial - put the summer fall / hiker map on one side of the map and the winter / spring skiing map on the other.

[Martin Gamache]

I've thought about that a lot - I've talked to quite a few hardcore backcountry folks and they seem to think that as long as they know what is glaciated, where the tree-line is located and what are the access roads, they are happy (they are not a fussy bunch - they just want to ski).  But I think you have a good point - a Landsat7 captured in the early spring would probably be beneficial - put the summer fall / hiker map on one side of the map and the winter / spring skiing map on the other.

<{POST_SNAPBACK}>

Would make more sense from a business P.O.V. to make it two maps as you can sell the same maps (sorta) twice for little additional work.

Did you perform the classification yourself?

Did you consider generating a new DEM from the TRIMM data?


As for your glacier boundary, you can probably even go with a dashed line....

How you gonna get this thing printed? 4, 5 or 6 colors? (something to keep in mind when picking those glacier isoline weights and color) Any printers in mind?

The whole thing looks real nice so far... post it when it's done....

mg

[Clark Geomatics]

Would make more sense from a business P.O.V. to make it two maps as you can sell the same maps (sorta) twice for little additional work. 

<{POST_SNAPBACK}>

I'm thinking that there would be quite a bit of extra work (and cost) as I would have to purchase and process an entirely separte scene captured during the early spring to get the snow line, winter features, etc.

Do you know of others who have created 'winter season' maps? With a winter scene, I guess one could keep the outlines of the glaciers for reference, however, the navigational aids on the map would be pretty sparse... not sure how that would go over with the skiing crowd.

Would love to see any examples - if anyone knows of any.

Did you perform the classification yourself?

<{POST_SNAPBACK}>

Yes, I do a lot of remote sensing work with various spectral satellite platforms (along with ASTER and the likes). The classification that you refer to is the result of a process that starts with ER Mapper and ends in Photoshop (for cleaning up extraneous pixels and to apply the land use tinting / textures). There's a few remote sensing tricks I use to get rid of the shadows and merge the Pan band, etc.

Did you consider generating a new DEM from the TRIMM data?

<{POST_SNAPBACK}>

Yes I did until I found out that there are massive intellectual property fees attached to using the TRIM data (that's over and above the $200 fee per 1:20k mapsheet - of which I need several). Strangely enough, the CDED data is based on TRIM and is in the public domain. Like we say in Canada - "No brainer, eh?"

How you gonna get this thing printed? 4, 5 or 6 colors? (something to keep in mind when picking those glacier isoline weights and color) Any printers in mind?

<{POST_SNAPBACK}>

This will be the first map I get printed / published - so, that said, I'm quite illiterate when it comes to colour separations, etc. Any advice here would be much appreciated. What I'm counting on is getting advice from a local map publisher who's put out tons of maps (ITMB). In terms of a printer, we've got a place called Hemlock Printers - supposedly higher-end stuff goes there - I'm still a bit in the dark.

I've uploaded a little map browser applet on our website that has a more saturated version of the land use classification without the isolines so you can get a better feeling for what the image looks like - hopefully the resolution is high enough to make sense.

Here it is: http://www.spatialvi...di/Template.htm

Cheers,

Jeff