Nathaniel

Forum Replies Created

Viewing 20 posts - 141 through 160 (of 206 total)
  • Author
    Posts
  • in reply to: Errors re-projecting 50m Ocean shapefile #4025

    Nathaniel
    Keymaster

    @Evzob: You can confirm this hypothesis by testing if the following shp reprojects:

    http://www.nacis.org/naturalearth/temp/50m_ocean_clipped.zip

    in reply to: Errors re-projecting 50m Ocean shapefile #4024

    Nathaniel
    Keymaster

    @Evzob: Short answer: because of a software limitation on your end, you’ll need to clip the shapefile to a +/- 180/90 bounding box. There is a small offset in the natural earth data outside of the normal extents (that is what “latitude or longitude exceeded limits” means). So find or make such a file (but don’t download the one off the NE site) and clip the oceans with it. Long answer: Commercial GIS software doesn’t care about minor issues like this, the open source mapping community needs to update their programs to deal with unexpected data extents.

    in reply to: NE on server? #4021

    Nathaniel
    Keymaster

    @Brandon: Bjørn has put together a simple browser for Natural Earth here: http://blog.thematicmapping.org/2010/01/natural-earth-browser.html

    There is also a simple Google Earth version out there.

    No one has make it available as a WFS/WCS. It would be great to have it up there in that form!

    Jill also points out a version that allows SHPs or at least vector (eps,pdf,svg) output of selected view would be a boon.

    If you figure out a solution we could ramp up no the horsepower, ideally with open source tools or ability to scale the $ version.

    _Nathaniel

    in reply to: Help using Natural Earth data in QGIS #4020

    Nathaniel
    Keymaster

    @TB: Sorry, but there is but that one geoTIFF at present. It was created to isolate a similar problem a couple months ago. I’d like to convert all the TIFFs to work in QGIS and other open-source GIS programs but that is several gigabytes to process. The onus also falls on the open-source community to support files that a GIS like ESRI’s ArcMap can read in fine (same goes for vector versions of Natural Earth that sometimes fail in Mapnik). In the case of the raster, the problem is likely that the TFW file has the correct info but the header in the TIF does not (or at least it’s encoded “wrong” to QGIS). There is an order of operations GIS apps use to pull out the coordinate system space on a TIFF image. Because it’s erroring on the TIF header (which is giving priority), QGIS never gets to the TFW version. One alternative for you is to open the TIF up in a normal image editor and save a copy out. That will strip the header from the new TIF and then QGIS should then read the TFW to locate the TIF image. You might have to tell QGIS that the projection is geographic WGS84, but the registration and cell size will be right in the TFW. Let me know if that works for you as a workaround.

    in reply to: Help using Natural Earth data in QGIS #4018

    Nathaniel
    Keymaster

    @TB: Please try the following version of the tif:

    http://www.nacis.org/naturalearth/temp/NE1_LR_LC_SR_GEOTIFF.zip

    Other people have reported it loaded better into open-source GIS systems.

    Reviewing your video, it looks like the coordinate system is fine on the image. What is wrong is QGIS is not reading the pixel size or upper-left registration point correctly. The extent should be ±180 ±90 and registered at -180, 90 (I’m not sure how QGIS measures the pixel cell offset so that could be slightly off from integer values). What it’s doing is assuming each pixel is equal to 1 decimal degree, and registereding the image at 0,0 (in the middle of the geographic coordiante space).

    I’m not seeing how to set the extent quickly in Google search. Please ask how to do that in the QGIS forums.

    in reply to: Help using Natural Earth data in QGIS #4016

    Nathaniel
    Keymaster

    @TB: That video is marked as private. Please reply here when it’s public.

    in reply to: Help using Natural Earth data in QGIS #4014

    Nathaniel
    Keymaster

    @TB: Sounds like QGIS is not paying attention to the projection info on the TIF file. You’ll need to use QGIS’s define coordinate reference system (CRS) tools. Looking at the QGIS manual, this can be changed by selecting your raster in the table of contents, getting the properties, and in the general tab, it will list a coordinate system (it might say NULL now) and there is a CHANGE button next to that. You want it to be something like: proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs

    in reply to: project raster NE II #4012

    Nathaniel
    Keymaster

    @Pstead: Glad you like the project, thanks! You can find the Natural Earth II raster imagery with water and drainage features burned into single file here: https://www.naturalearthdata.com/downloads/10m-natural-earth-2/10m-natural-earth-2-with-shaded-relief-water-and-drainages/. It looks like that file has a small naming bug, though. Rename the .twf.txt so it looses the .txt part leaving just “NE2_LR_LC_SR_W_DR.tfw”. After you rename it, then import into ArcMap or you GIS of choice. You’ll need to define the projection, and then you can reproject.

    Searching on Google for “set dataframe projection arcmap” will turn up some good tutorials on defining a projection (in this case Geographic WGS84) and reprojecting in ArcMap and using Toolbox.

    Note: because of some ArcMap weirdness, you may need to relaunch your ArcMap instance and NOT calculate image pyramids to see the projection change with the imagery.


    Nathaniel
    Keymaster

    @MSM: The wording was wonky, I’ve changed it. The 10m raster and vector register precisely with each other, as do the 50m set.

    in reply to: Reduce file size #4008

    Nathaniel
    Keymaster

    Open the GeoTIFF in an image editor and downsize by 50% for the 50m small res files to get a file under 100 mb. Then in your mapping app import and, since the TFW world file is now out of date, set the image extents back to ±180, ±90.


    Nathaniel
    Keymaster
    in reply to: river tapers via stroke weight attribute #4009

    Nathaniel
    Keymaster

    @WWallace: Please download the 1.1.1 update of Natural Earth here to get the fixed river tapers: https://www.naturalearthdata.com/blog/miscellaneous/version-1-1-release-notes/

    The 1.0 release had the strokeWeight field set to text instead of number for TWO files: the “scaleRank” version and the dissolved rivers version. With the new version of scaleRank file you’ll be able to set a quantitative (numerical) legend based on that field. Do graduated line weights from symbolization range of .25 to 8 point weight. You should see very large rivers like the Amazon, Congo, and Mississippi have large outflows on the coast and the headwaters of all rivers shrink up.

    Because the scaleRank file has many segments, it’s not good for labeling. Add the dissolved file (provided) and set it’s symbolization to no fill, no stroke, but label true using the name field. You can also do a selection querry so only rivers are labeled, not the parts of rivers in lakes (feature class = lake centerline).

    I hope to push out the rivers and other updated files into the main branch (downloads on the site) soon. Please subscribe to the update email list to stay current with Natural Earth.


    Nathaniel
    Keymaster

    @WWallace: I’m not seeing the same thing on my end in Arc 9.3. What happens when you import other Natural Earth tiffs? What happens if you open the tiff up in an image viewer outside of ArcMap, are the colors right then?

    in reply to: What software to buy for map enhancements #4003

    Nathaniel
    Keymaster

    If you’re dealing with vectors in the GIS, Illustrator would be your best bet for further edits. Photoshop can do live-vector objects, but it is not made for that. The app has color enhancement tools, and if you want more Photoshop color transforms, you can get the Phantasm range of plugins from http://www.astutegraphics.com/. Google Earth Pro (with SHP import) can do almost all of what Bryce can do but with real GIS data rather than hack workarounds. Illustrator has week 3d, true.

    in reply to: Thematic codes #4000

    Nathaniel
    Keymaster

    The website Statoids has a good overview of country code systems:

    http://www.statoids.com/wab.html

    The following 4 columns are unique to the Natural Earth administrative-0 coding system, but somewhat follow the ISO 3 digit alpha codes: sov_a3, adm_a3, gu_a3, su_a3

    The “map_color” column derives from a US State Department political world map where no neighboring country has the same color, but all territories of a sovereign country share the same color.

    The columns COG to Pays_region are from the French statistical authority.

    For the cShapes (Correlates of War) link column, use “gw_id” which conforms to Gleditsch and Ward’s modified system. The website with the GIS shapefiles and further info: http://nils.weidmann.ws/projects/cshapes

    Also, since some countries include a “The” beginning their name, the “SortName” provides a way of listing ignoring that prefix.

    The people (population) and gdp_usd (gross domestic product, United States dollars) columns are mostly from 2009 and from World Bank, UN, etc. But where those bodies do not parse administrative countries using the same formula as Natural Earth, they have been supplemented with other sources. These columns are provided to give rough estimate of these values to determine relative rankings between countries, not to be an authoritative source for a particular country.

    in reply to: Sub-countries of France vs. other UK/US/Australia #3998

    Nathaniel
    Keymaster

    While they would get the adm0 name “France”, you could label them with the geoUnit name. The preferred method is to label them: “GeoUnit (Terr.)”. Example: Terr. = Fr. and GeoUnit = Guadeloupe. So test if GeoUnit == the Adm0 Name then output = Adm0 Name else output = GeoUnit & ” (” & “Adm0 Name & “)” endif. This is slightly easier (via data column attributes) in the forthcoming 1.1 update.

    in reply to: Sub-countries of France vs. other UK/US/Australia #3996

    Nathaniel
    Keymaster

    Overseas regions of France like:

    1. Guadeloupe (in the Caribbean)

    2. French Guiana (in South America)

    3. Martinique (in the Caribbean)

    4. Réunion (in the Indian Ocean)

    are considered geo-units of France in the Natural Earth coding system. They are like Alaska and Hawaii in USA, integral parts of France. While the ISO considers them “countries” they are not. France disagrees with the ISO coding but has not logged an official complaint. Other units (territories, collectives) of France are listed at the adm-0 “country” level in Natural Earth.

    http://simple.wikipedia.org/wiki/Overseas_Regions_of_France

    Does that help?

    in reply to: boundary-lines-land vs. -maritime-indicator #3994

    Nathaniel
    Keymaster

    @Jochen: think of the land boundary lines being inclusive of coastal waters, so sometimes you’ll find them in inlets and in straights or bays. In many atlases you’d see these drawn as land boundaries, not water boundaries. For the admin-1 lines and polys, there is more of an issue as they purposely go thru larger lakes. I agree it would be good to have a subclass for each. It’d be nice to push that change out in the next update, but it may need to wait for 1.2 update.

    in reply to: Natural Earth data BitTorrent mirror #3993

    Nathaniel
    Keymaster

    Your welcome, Mike! I’ll follow up with you as updated files are pushed out.

    in reply to: gridded (raster?) country information #3992

    Nathaniel
    Keymaster

    @wxmangjh: A prebuilt raster of the admin-0 countries in Natural Earth is not available as a default download. However, it should be simple for the vector files to be converted to that format. Good luck!

Viewing 20 posts - 141 through 160 (of 206 total)