Compass Rose Orientation

[Nexis]

A basic cartographic question. How do I orient my compass rose? When I look at my map (which is just south of the equator) all the longitudes are curving to meet at the south pole. In my addled brain it seems to me I would orient the compass to follow one of the lines as it curves to the north. I'm certain I am totally wrong but it's something I can't seem to find any info on the webz.

[waldronate]

It depends on your cultural preferences for the map. Some cultures prefer South at the top of the map, some North, some East, etc. A modern map such as from a national mapping agency will often have a north marker that incorporates a true north and a magnetic north marker of some sort.

If the goal is to mark true north then your compass would align to the direction of true north at its location as you have surmised.

[töff]

If you're concerned about the rose being "straight" and not really aligning meaningfully to the curvature of the lines of longitude, you can always warp the rose to match.

Why not post what you have? I'm SURE we'd all love to see it! (It's why we're here!)

[Robbie]

I would think that if your map has lines on it that curve, then those lines are your markers and you wouldn't need a rose per se, as much as just a north pointer that is near a line.

[Nexis]

If you're concerned about the rose being "straight" and not really aligning meaningfully to the curvature of the lines of longitude, you can always warp the rose to match.

Why not post what you have? I'm SURE we'd all love to see it! (It's why we're here!)

OK this is what I have.

[Hai-Etlik]

First, A rose is only for maps where the directions do not vary. That means either Mercator projection, or relatively large scale (Covering small areas).

If a line of constant bearing (Rhumb Line or Loxodrome) is not straight on your map, you can't use a rose.

A simple north arrow doesn't imply as much precision, but I'd still be leery about it. The graticule (The grid) by itself should be enough in most cases.

One thing that might work is to extend a meridian from the graticule and put an arrow on that to indicate north. Of course this depends on an edge at the north of the map. It would work for Winkel Tripel, or Lambert Conformal Conic, but not Polar Stereographic.

[Nexis]

First, A rose is only for maps where the directions do not vary. That means either Mercator projection, or relatively large scale (Covering small areas).

If a line of constant bearing (Rhumb Line or Loxodrome) is not straight on your map, you can't use a rose.

A simple north arrow doesn't imply as much precision, but I'd still be leery about it. The graticule (The grid) by itself should be enough in most cases.

One thing that might work is to extend a meridian from the graticule and put an arrow on that to indicate north. Of course this depends on an edge at the north of the map. It would work for Winkel Tripel, or Lambert Conformal Conic, but not Polar Stereographic.

How about Lambert Azimuthal Equal-Area?

But so if I have my maps displaying a well marked Graticule I can use the rose as mostly decorative and not worry about direction?

[Hai-Etlik]

How about Lambert Azimuthal Equal-Area?

But so if I have my maps displaying a well marked Graticule I can use the rose as mostly decorative and not worry about direction?

If you have a rose, it should be correct. If it can't be correct, you shouldn't have one at all. In fact you should avoid anything that might even be mistaken for one. Roses tend to be confined to navigational maps Thematic and reference maps don't normally have them.

Simple north arrows have looser tolerances but they are still not "purely decorative".

LAEA would depend on where the point of tangency is and the extent of the map, if the north pole is part of the map, my idea wouldn't work. If the scale is large enough and you are close to the point of tangency, the distortion will be negligible and you could use a rose safely.

Scale bars are similar, if the bar is not effectively correct for the whole map, you shouldn't have one. A global map should NEVER have a scale bar as it simply can't be correct for the whole map.

PS: A slight correction, Azimuthal Equidistant maps can justifiably have a scale bar even for very small scales but it only applies to distances along lines through the point of tangency, and this should be clearly indicated.

[Nexis]

If you have a rose, it should be correct. If it can't be correct, you shouldn't have one at all. In fact you should avoid anything that might even be mistaken for one. Roses tend to be confined to navigational maps Thematic and reference maps don't normally have them.

Simple north arrows have looser tolerances but they are still not "purely decorative".

LAEA would depend on where the point of tangency is and the extent of the map, if the north pole is part of the map, my idea wouldn't work. If the scale is large enough and you are close to the point of tangency, the distortion will be negligible and you could use a rose safely.

Scale bars are similar, if the bar is not effectively correct for the whole map, you shouldn't have one. A global map should NEVER have a scale bar as it simply can't be correct for the whole map.

PS: A slight correction, Azimuthal Equidistant maps can justifiably have a scale bar even for very small scales but it only applies to distances along lines through the point of tangency, and this should be clearly indicated.

OK so rose is out. Then I should be using a kartush for the map info.

"along lines through the point of tangency" Hmmmm...Nope lost me there

[Hai-Etlik]

"along lines through the point of tangency" Hmmmm...Nope lost me there

Azimuthal projections are based on planes that just touch the globe. In math we call that "being tangent" and the point they touch is the "point of tangency"

Such projections get more distorted the further you get from that point. The details of the distortion depend on the projection.