Changes between Version 1 and Version 2 of FAQ
- Timestamp:
- 05/15/08 21:27:07 (5 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
FAQ
v1 v2 33 33 A default build and install on Unix will normally build knowledge of the directory where the grid shift files are installed into the PROJ.4 library (usually /usr/local/share/proj). On Windows the library is normally built thinking that C:\PROJ\NAD is the installed directory for the grid shift files. If the built in concept of the PROJ.4 data directory is incorrect, the PROJ_LIB environment can be defined with the correct directory. 34 34 35 <h2><a name="nad2783_debug">How do I debug problems with NAD27/NAD83 datum shifting?</a></h2> 35 == How do I debug problems with NAD27/NAD83 datum shifting? == 36 36 37 <ol> 38 <li> Verify that you have the binary files (eg. /usr/local/share/proj/conus) 39 installed on your system. If not, see the previous question. 40 <p> 41 <li> Try a datum shifting operation in relative isolation, such as with the 42 cs2cs command listed above. Do you get reasonable results? If not it is 43 likely the grid shift files aren't being found. Perhaps you need to define 44 PROJ_LIB? <p> 37 * Verify that you have the binary files (eg. /usr/local/share/proj/conus) installed on your system. If not, see the previous question. 38 * Try a datum shifting operation in relative isolation, such as with the cs2cs command listed above. Do you get reasonable results? If not it is likely the grid shift files aren't being found. Perhaps you need to define PROJ_LIB? 39 * The cs2cs command and the underlying pj_transform() API know how to do a grid shift as part of a more complex coordinate transformation; however, it is imperative that both the source and destination coordinate system be defined with appropriate datum information. That means that implicitly or explicitly there must be a +datum= clause, a +nadgrids= clause or a +towgs84= clause. For instance "cs2cs +proj=latlong +datum=NAD27 +to +proj=latlong +ellps=WGS84" won't work because defining the output coordinate system as using the ellipse WGS84 isn't the same as defining it to use the datum WGS84 (use +datum=WGS84). If either the input or output are not identified as having a datum, the datum shifting (and ellipsoid change) step is just quietly skipped! 40 * The PROJ_DEBUG environment can be defined (any value) to force extra output from the PROJ.4 library to stderr (the text console normally) with information on what data files are being opened and in some cases why a transformation fails. Note that PROJ_DEBUG support is not yet very mature in the PROJ.4 library. 41 * The "-v" flag to cs2cs can be useful in establishing more detail on what parameters being used internally for a coordinate system. This will include expanding the definition of +datum clause. 45 42 46 <li> The cs2cs command and the underlying pj_transform() API know how to 47 do a grid shift as part of a more complex coordinate transformation; however, 48 it is imperative that both the source and destination coordinate system be 49 defined with appropriate datum information. That means that implicitly or 50 explicitly there must be a +datum= clause, a +nadgrids= clause or 51 a +towgs84= clause. For instance "cs2cs +proj=latlong +datum=NAD27 +to 52 +proj=latlong +ellps=WGS84" won't work because defining the output coordinate 53 system as using the ellipse WGS84 isn't the same as defining it to use the 54 datum WGS84 (use +datum=WGS84). If either the input or output are not 55 identified as having a datum, the datum shifting (and ellipsoid change) step 56 is just quietly skipped!<p> 43 == How do I use EPSG coordinate system codes with PROJ.4? == 57 44 58 <li> The PROJ_DEBUG environment can be defined (any value) to force extra 59 output from 60 the PROJ.4 library to stderr (the text console normally) with information on 61 what data files are being opened and in some cases why a transformation fails. 62 Note that PROJ_DEBUG support is not yet very mature in the PROJ.4 library.<p> 45 There is somewhat imperfect translation between 2d geographic and projected coordinate system codes and PROJ.4 descriptions of the coordinate system available in the epsg definition file that normally lives in the proj/nad directory. If installed (it is installed by default on Unix), it is possible to use EPSG numbers like this: 63 46 64 <li> The "-v" flag to cs2cs can be useful in establishing more detail on 65 what parameters being used internally for a coordinate system. This will 66 include expanding the definition of +datum clause. <p> 67 68 </ol> 69 70 <!--------------------------------------------------------------------------> 71 72 <h2><a name="epsg">How do I use EPSG coordinate system codes with PROJ.4?</a></h2> 73 74 There is somewhat imperfect translation between 2d geographic and 75 projected coordinate system codes and PROJ.4 descriptions of the 76 coordinate system available in the <tt>epsg</tt> definition file that 77 normally lives in the <tt>proj/nad</tt> directory. If installed (it is 78 installed by default on Unix), it is possible to use EPSG numbers like this: 79 80 <pre> 47 {{{ 81 48 % cs2cs -v +init=epsg:26711 82 49 # ---- From Coordinate System ---- … … 92 59 # 93 60 # +proj=latlong +datum=NAD27 +ellps=clrk66 +nadgrids=conus,ntv1_can.dat 94 </pre> 61 }}} 95 62 96 The <tt>proj/nad/epsg</tt> file can be browsed and searched in a text editor 97 for coordinate systems. There are known to be problems with some coordinate 98 systems, and any coordinate systems with odd axes, a non-greenwich prime 99 meridian or other quirkyness are unlikely to work properly. Caveat Emptor!<p> 63 The proj/nad/epsg file can be browsed and searched in a text editor for coordinate systems. There are known to be problems with some coordinate systems, and any coordinate systems with odd axes, a non-greenwich prime meridian or other quirkyness are unlikely to work properly. Caveat Emptor! 100 64 101 <!--------------------------------------------------------------------------> 65 == How do I use 3 parameter and 7 parameter datum shifting == 102 66 103 <h2><a name="datum_37">How do I use 3 parameter and 7 parameter datum shifting</a></h2> 67 Datum shifts can be approximated with 3 and 7 parameter transformations. Their use is more fully described in the <a href="gen_parms.html#towgs84">towgs84</a> discussions. 104 68 105 Datum shifts can be approximated with 3 and 7 parameter transformations. 106 Their use is more fully described in the 107 <a href="gen_parms.html#towgs84">towgs84</a> discussions.<p> 69 == Does PROJ.4 work in different international numeric locales? == 108 70 71 No. PROJ.4 makes extensive use of sprintf() and atof() internally to translate numeric values. If a locale is in effect that modifies formatting of numbers, altering the role of commas and periods in numbers, then PROJ.4 will not work. This problem is common in some European locales. 109 72 110 <!--------------------------------------------------------------------------> 111 112 <h2><a name="locale">Does PROJ.4 work in different international numeric locales?</a></h2> 113 114 No. PROJ.4 makes extensive use of sprintf() and atof() internally to translate 115 numeric values. If a locale is in effect that modifies formatting of 116 numbers, altering the role of commas and periods in numbers, then PROJ.4 117 will not work. This problem is common in some European locales.<p> 118 119 On unix-like platforms, this problem can be avoided by forcing the use 120 of the default numeric locale by setting the LC_NUMERIC environment variable 121 to C. <p> 73 On unix-like platforms, this problem can be avoided by forcing the use of the default numeric locale by setting the LC_NUMERIC environment variable to C. 122 74 123 75 eg. 124 <pre> 76 {{{ 125 77 $ export LC_NUMERIC=C 126 78 $ proj ... 127 </pre> 79 }}} 128 80 129 <!--------------------------------------------------------------------------> 81 == Changing Ellipsoid / Why can't I convert from WGS84 to Virtual Earth Mercator? == 130 82 131 <h2><a name="sphere_as_wgs84">Changing Ellipsoid / Why can't I convert from WGS84 to Virtual Earth Mercator?</a></h2> 83 The coordinate system definition for Virtual Earth Mercator is as follows, which uses a sphere as the earth model for the mercator projection. 132 84 133 The coordinate system definition for Virtual Earth Mercator is as follows, 134 which uses a sphere as the earth model for the mercator projection. 135 136 <pre> 85 {{{ 137 86 +proj=merc +a=6378137 +b=6378137 +lat_ts=0.0 +lon_0=0.0 138 87 +x_0=0.0 +y_0=0 +k=1.0 +units=m +no_defs 139 </pre> 88 }}} 140 89 141 90 But, if you do something like: 142 91 143 <pre> 92 {{{ 144 93 cs2cs +proj=latlong +datum=WGS84 145 94 +to +proj=merc +a=6378137 +b=6378137 +lat_ts=0.0 +lon_0=0.0 146 95 +x_0=0.0 +y_0=0 +k=1.0 +units=m +no_defs 147 </pre> 96 }}} 148 97 149 to convert between WGS84 and mercator on the sphere there will be substantial 150 shifts in the Y mercator coordinates. This is because internally cs2cs is 151 having to adjust the lat/long coordinates from being on the sphere to being 152 on the WGS84 datum which has a quite differently shaped ellipsoid.<p> 98 to convert between WGS84 and mercator on the sphere there will be substantial shifts in the Y mercator coordinates. This is because internally cs2cs is having to adjust the lat/long coordinates from being on the sphere to being on the WGS84 datum which has a quite differently shaped ellipsoid. 153 99 154 In this case, and many other cases using spherical projections, the desired 155 approach is to actually treat the lat/long locations on the sphere as if 156 they were on WGS84 without any adjustments when using them for converting 157 to other coordinate systems. The solution is to "trick" PROJ.4 into applying 158 no change to the lat/long values when going to (and through) WGS84. This 159 can be accomplished by asking PROJ to use a <i>null</i> grid shift file 160 for switching from your spherical lat/long coordinates to WGS84.<p> 100 In this case, and many other cases using spherical projections, the desired approach is to actually treat the lat/long locations on the sphere as if they were on WGS84 without any adjustments when using them for converting to other coordinate systems. The solution is to "trick" PROJ.4 into applying no change to the lat/long values when going to (and through) WGS84. This can be accomplished by asking PROJ to use a null grid shift file for switching from your spherical lat/long coordinates to WGS84. 161 101 162 <pre> 102 {{{ 163 103 cs2cs +proj=latlong +datum=WGS84 164 104 +to +proj=merc +a=6378137 +b=6378137 +lat_ts=0.0 +lon_0=0.0 165 105 +x_0=0.0 +y_0=0 +k=1.0 +units=m +nadgrids=@null +no_defs 166 </pre> 106 }}} 167 107 168 Note the strategic addition of <tt>+nadgrids=@null</tt> to the spherical 169 projection definition.<p> 108 Note the strategic addition of +nadgrids=@null to the spherical projection definition. 170 109 171 Similar issues apply with many other datasets distributed with projections 172 based on a spherical earth model - such as many NASA datasets, and also 173 (I think) the Google Maps mercator projection.<p> 110 Similar issues apply with many other datasets distributed with projections based on a spherical earth model - such as many NASA datasets, and also (I think) the Google Maps mercator projection. 174 111 175 <!-------------------------------------------------------------------------->176 177 <hr>178 179 Requests to add items to the frequently asked questions list180 <a href="http://bugzilla.remotesensing.org/enter_bug.cgi?product=PROJ.4">181 can be entered</a> in bugzilla.<p>182 </body>183 </html>
