mapug.html

m_map是加拿大学者编写的一个在matlab上绘地图的软件

coastlines or other awkwardly shaped or aligned regions. It is
conformal but not equal area. The following properties govern this
projection: </a>
    <h4><a name="p2.2"> <code> &lt;,'lon&lt;gitude&gt;',[ G1 G2 ]&gt; <br>
&lt;,'lat&lt;itude&gt;', [ L1 L2 ]&gt; </code> </a></h4>
    <a name="p2.2">Two points specify a great circle, and thus the
limits of this map (it is assumed that the region near the shortest of
the two arcs is desired). The 2 points (G1,L1) and (G2,L2) are thus at
the center of either the top/bottom or left/right sides of the map
(depending on the <code>'direction'</code> property). </a>
    <h4><a name="p2.2"> <code> &lt;,'asp&lt;ect&gt;',value&gt; </code>
    </a></h4>
    <a name="p2.2"> This specifies the size of the map in the direction
perpendicular to the great circle of tangency, as a proportion of the
length shown. An aspect ratio of 1 results in a square map, smaller
numbers
result in skinnier maps. Aspect ratios &gt;1 are possible, but not
recommended. </a>
    <h4><a name="p2.2"> <code> &lt;,'dir&lt;ection&gt;',( 'horizontal'
| 'vertical' ) </code> </a></h4>
    <a name="p2.2"> This specifies whether the great circle of tangency
will be horizontal on the page (for making short wide maps), or
vertical (for tall thin maps). </a>
    <h4><li><a name="p2.2"> Transverse Mercator </a></li>
    </h4>
    <a name="p2.2">The Transverse Mercator is a special case of the
oblique mercator when the great circle of tangency lies along a
meridian of longitude, and is therefore conformal. It is often used for
large-scale maps and
charts. The following properties govern this projection: </a>
    <h4><a name="p2.2"> <code> &lt;,'lon&lt;gitude&gt;',[min max]&gt; <br>
&lt;,'lat&lt;itude&gt;',[min max]&gt; </code> </a></h4>
    <a name="p2.2"> These specify the limits of the map. </a>
    <h4><a name="p2.2"> <code> &lt;,'clo&lt;ngitude&gt;',value&gt; </code>
    </a></h4>
    <a name="p2.2"> Although it makes most sense in general
to specify the central meridian as the meridian of tangency (this is
the
default), certain map classification systems (noteably UTM) use only a
fixed set of central longitudes, which may not be in the map center. </a>
    <h4><a name="p2.2"> <code> &lt;,'rec&lt;tbox&gt;', ( 'on' | 'off'
)&gt; </code> </a></h4>
    <a name="p2.2">The map limits can either be based on
latitude/longitude (the default), or the map boundaries can form an
exact rectangle. The
difference is small for large-scale maps. Note: Although this
projection
is similar to the Universal Transverse Mercator (UTM) projection, the
latter is actually ellipsoidal in nature. </a>
    <h4><li><a name="p2.2"> Universal Transverse Mercator (UTM) </a></li>
    </h4>
    <a name="p2.2">UTM maps are needed only for high-quality maps of
small regions of the globe (less than a few degrees in longitude). This
is an ellipsoidal projection. Options are similar to those of the
Transverse
Mercator, with the addition of </a>
    <h4><a name="p2.2"> <code> &lt;,'zon&lt;e&gt;', value 1-60&gt; </code>
    </a></h4>
    <a name="p2.2"> </a>
    <h4><a name="p2.2"> <code> &lt;,'hem&lt;isphere&gt;',value
0=N,1=S&gt; </code> </a></h4>
    <a name="p2.2"> These are computed automatically if not specified.
The ellipsoid defaults to <code>'normal'</code>, a spherical earth of
radius 1 unit, but other options can also be chosen using the following
property: </a>
    <h4><a name="p2.2"> <code> &lt;,'ell&lt;ipsoid&gt;', ellipsoid&gt;
    </code> </a></h4>
    <a name="p2.2"> For a list of available ellipsoids try <code>m_proj('get','utm')</code>.
    </a>
    <p> <a name="p2.2">The big difference between UTM and all the
other projections is that for ellipsoids other than <code>'normal'</code>
the projection
coordinates are in meters of easting and northing. To take full
advantage
of this it is often useful to call <code>m_proj</code> with <code>'rectbox'</code>
set to <code>'on'</code> and not to use the long/lat grid generated
by <code>m_grid</code> (since the regular matlab grid will be in units
of meters). </a></p>
    <h4><li><a name="p2.2"> Sinusoidal </a></li>
    </h4>
    <a name="p2.2">This projection is usually called
"pseudo-cylindrical" since parallels of latitude appear as straight
lines, similar to their appearance in cylindrical projections tangent
to the equator. However, meridians curve together in this projection in
a sinusoidal way (hence the name), making this map equal-area. </a>
    <h4><li><a name="p2.2"> Gall-Peters </a></li>
    </h4>
    <a name="p2.2">Parallels of latitude and meridians both appear as
straight lines, but the vertical scale is distorted so that area is
preserved. This is useful for tropical areas, but the distortion in
polar areas is extreme. </a>
  </ol>
  <h3><a name="p2.2"> <li> <br>
  </li>
  </a><a name="p2.3"> Conic Projections </a></h3>
  <a name="p2.3">Conic projections result from projecting onto a cone
wrapped around the sphere. The vertex of the cone lies on the
rotational axis of the sphere. The cone is either tangent at a single
latitude, or can intersect the sphere at two separated latitudes. It is
a useful projection for mid-latitude areas of large east-west extent.
The following properties affect these projections: </a>
  <h4><a name="p2.3"> <code> &lt;,'lon&lt;gitude&gt;',[min max]&gt; <br>
&lt;,'lat&lt;itude&gt;',[min max]&gt; </code> </a></h4>
  <a name="p2.3"> These specify the limits of the map. </a>
  <h4><a name="p2.3"> <code> &lt;,'clo&lt;ngitude&gt;',value&gt; </code>
  </a></h4>
  <a name="p2.3"> The central longitude appears as a vertical on the
page. The default value is the mean longitude, although it may be set
to any value (even one outside the limits). </a>
  <h4><a name="p2.3"> <code> &lt;,'par&lt;allels&gt;',[lat1 lat2]&gt; </code>
  </a></h4>
  <a name="p2.3">The standard parallels can be specified. Either one or
two parallels can be given, the default is a single parallel at the
mean latitude </a>
  <h4><a name="p2.3"> <code> &lt;,'rec&lt;tbox&gt;', ( 'on' | 'off'
)&gt; </code> </a></h4>
  <a name="p2.3">The map limits can either be based on
latitude/longitude (the default), or the map boundaries can form an
exact rectangle which
contain the given limits. Unless the region being mapped is small, it
is
best to leave this <code> 'off' </code>. </a>
  <ol>
    <a name="p2.3"> </a>
    <h4><li><a name="p2.3"> Albers Equal-Area Conic </a></li>
    </h4>
    <a name="p2.3">This projection is equal-area, but not conformal </a>
    <h4><li><a name="p2.3"> Lambert Conformal Conic </a></li>
    </h4>
    <a name="p2.3">This projection is conformal, but not equal-area. </a>
  </ol>
  <h3><a name="p2.3"> <li> <br>
  </li>
  </a><a name="p2.4"> Miscellaneous global projections </a></h3>
  <a name="p2.4">There are a number of projections which don't really
fit into any of the above categories. Mostly these are global
projections (i.e. they show the whole world), and they have been
designed to be "pleasing to the eye". I'm not sure what use they are in
general, but they make
nice logos! </a>
  <ol>
    <a name="p2.4"> </a>
    <h4><li><a name="p2.4"> Hammer-Aitoff </a></li>
    </h4>
    <a name="p2.4">An equal-area projection with curved meridians and
parallels. </a>
    <h4><li><a name="p2.4"> Mollweide </a></li>
    </h4>
    <a name="p2.4"> Also called the Elliptical or Homolographic
Equal-Area Projection. Parallels are straight (and parallel) in this
projection.
Note that </a><a href="../map.html#e4">example 4</a> shows a rather
sophisticated use designed to reduce distortion, a more standard map
can be made using
    <pre>m_proj('mollweide');<br>m_coast('patch','r');<br>m_grid('xaxislocation','middle');<br></pre>
    <h4><li><a name="p2.4"> Robinson </a></li>
    </h4>
    <a name="p2.4"> Not equal-area OR conformal, but supposedly "pleasing to the eye".
  </ol>
  <h3> <li> <a name="p2.5"> Yeah, but which projection should I use?</a></li>
  </h3>
  <a name="p2.5">Well, it depends really on how large an area you are
mapping. Usually, maps of the whole world are Mercator, although often
the Miller Cylindrical projection looks better because it doesn't
emphasize the polar areas as much. Another choice is the Hammer-Aitoff
or Mollweide (which has meridians curving together near the poles).
Both are equal-area. It's probably not a good idea to use these
projections for maps that don't have the equator somewhere near the
middle. The Robinson projection is not equal-area or conformal, but was the choice of National Geographic (for a while, anyway),
and also appears in the IPCC reports.</a>
  <p> <a name="p2.5">If you are plotting something with a large
north/south extent, but not very wide (say, North and South America, or
the North
and South Atlantic), then the Sinusoidal or Mollweide projections will
look pretty good. Another choice is the Transverse Mercator, although
that
is usually used only for very large-scale maps. </a></p>
  <p><a name="p2.5"> For smaller areas within one hemisphere or other
(say, Australia, the United States, the Mediterranean, the North
Atlantic) you might pick a conic projection. The differences between
the two available conic projections are subtle, and if you don't know
much about projections it probably won't make much difference which one
you use. </a></p>
  <p> <a name="p2.5">If you get smaller than that, it doesn't matter a
whole lot which projection you use. One projection I find useful in
many cases is the Oblique Mercator, since you can align it along a long
(but narrow) coastal area. If map limits along lines of
longitude/latitude are OK, use a Transverse Mercator or Conic
Projection. The UTM projection is also useful. </a></p>
  <p> <a name="p2.5">Polar areas are traditionally mapped using a
Stereographic projection, since for some reason it looks nice to have a
"bullseye" pattern of latitude lines. </a></p>
  <p> <a name="p2.5">If you want to get a quick idea of what any
projection looks like, default parameters for all functions are set for
a "typical" usage, i.e. to get a quick idea of what any projection
looks like, you
can do so without having to figure out a lot of numerical values: </a></p>
  <pre><a name="p2.5">m_proj('stereographic');  % Example for stereographic projection<br>m_coast;<br>m_grid;<br></a></pre>
  <p> </p>
  <li>
    <h3><a name="p2.6">Map scales</a></h3>
  </li>
  <a name="p2.6">M_Map usually scales the map so that it fits exactly
within the current axes. If you just want a nice picture (which is
mostly the case) then this is exactly what you need. On the other hand,
sometimes you want to print things out at some exact scale (i.e. if you
really much prefer sitting at your desk with a ruler and a piece of
paper trying to figure out how far apart Bangkok and Tokyo are). Use
the <code>m_scale</code> primitive for this - for a 1:250000 map, call
  </a>
  <pre><a name="p2.6">m_scale(250000);<br></a></pre>
  <a name="p2.6">after you have drawn everything (Be careful - a
1:250000 map of the world is a lot bigger than 8.5"x11" sheet of
paper). </a>
  <p> <a name="p2.6">This option is usually only useful for
large-scale maps, i.e. maps of very small areas). </a></p>
  <p> <a name="p2.6">If you wish to know the current scale, calling <code>m_scale</code>
without any parameters will calculate and return that value. </a></p>
  <p><a name="p2.6"> To return to the default scaling call <code>m_scale('auto')</code>.
  </a></p>
  <p> <a name="p2.6">(PS - If you do want to find distances from
Bangkok
to anywhere, plot an azimuthal equidistant projection of the world
centered on Bangkok (13 44'N, 100 30'E), and choose a fairly small
scale, like
1:200,000,000). Another option would be to use range rings, see </a><a
 href="../map.html#e11">example 11</a>.</p>
  <h3> <li> <a name="p2.7"> </a> <a name="p2.7">Map coordinate
systems - geographic and geomagnetic.</a></li>
  </h3>
  <a name="p2.7"> </a>
  <p><a name="p2.7"> </a><a name="p2.6">Latitude/Longitude is the
usual coordinate system for maps. In some cases UTM coords are also
used, but these are really just a simple transformation based on the
location of the equator and certain lines of longitude. On the other
hand, there are occasions when a coordinate system based on some other
set of axes is useful. For example, in space
physics data is often projected in coordinates based on the magnetic
poles.
&nbsp;M_Map has a limited capabality to deal with data in these other
coordinate
systems. m_coord allows you to chnage the coordinate system from
geographic
to geomagnetic.&nbsp; The following code gives you the idea:<br>
  </a></p>
  <p><a name="p2.6"><code>&nbsp;lat=[25*ones(1,100) 50*ones(1,100) 25];<br>
lon=[-99:0 0:-1:-99 -99];<br>
  <br>
clf<br>
subplot(121);<br>
m_coord('IGRF2000-geomagnetic'); % Treat all lat/longs as geomagnetic<br>
m_proj('stereographic');<br>
m_coast;<br>
m_grid;<br>
m_line(lon,lat,'color','r');&nbsp;&nbsp;&nbsp;&nbsp; % "lat/ln" assumed