I want to export a series of Landsat scenes in Africa (that natively have different UTM projections) and work on them in GIS software. To have the pixel boundaries at overlapping regions properly, I think I have to reproject my scenes to a common projection, for which I would like to use Africa_Albers_Equal_Area_Conic projection. I find this projection with authority ESRI:102022 on spatialreference.org, with below wkt definition:
PROJCS["Africa_Albers_Equal_Area_Conic",
GEOGCS["GCS_WGS_1984",
DATUM["WGS_1984",
SPHEROID["WGS_1984",6378137,298.257223563]],
PRIMEM["Greenwich",0],
UNIT["Degree",0.017453292519943295]],
PROJECTION["Albers_Conic_Equal_Area"],
PARAMETER["False_Easting",0],
PARAMETER["False_Northing",0],
PARAMETER["longitude_of_center",25],
PARAMETER["Standard_Parallel_1",20],
PARAMETER["Standard_Parallel_2",-23],
PARAMETER["latitude_of_center",0],
UNIT["Meter",1],
AUTHORITY["EPSG","102022"]]
I put this as the argument of ee.Projection() function and do the export this way (imageA is the input image):
var EPSG102022 = ee.Projection('PROJCS["Africa_Albers_Equal_Area_Conic",...)
var imageA_2 = imageA.reproject({crs:EPSG102022, scale:30})
var projectionA_2 = imageA_2.projection().getInfo()
Export.image.toDrive({
image: imageA_2,
description: 'TEST_A_reprojectWscale',
crs: projectionA_2.wkt,
scale: 30,
region: ROI
});
But when I do the export and load the original and reprojected image on ArcGIS, I see the reprojected image being misplaced. When I look at the reprojected image spatial reference in ArcGIS, I see that there is no "longitude_of_center" parameter there, brt there is a parameter named "central meridian" which is set to zero. Also the authority is not shown and it only says the projection authority is "custom".
Then I renamed "longitude_of_center" in the wkt string to "central_meridian" and exported the image again. The image was still misplaced. When I looked at its spatial reference, I found that the central meridian parameter is still zero, although I set it to be 25 in the wkt definition.
Then I set the "central_meridian" parameter in wkt definition to zero and exported the images again. This time the image was correctly placed on the map.
I am completely confused. Why there are different names for the projection parameters in different wkt's and why different softwares name the projection definition parameters differently? This looks like a very poor standard. Also, why ArcGIS does not follow the definition that I put in the exported file (setting central meridian to 25)?
Related
I am trying to map a geojson file (A map of Alaska precincts, downloaded from the division of elections and then converted online to geojson) onto a choropleth map using folium, the problem is the coordinates are in 7-digit numbers like this:
[ -16624764.227, 8465801.1497 ]
I read on a similar post that this was most likely a US coordinate system like UTM or State Plane, and recommended using an API to reproject it. Is it also possible to access the coordinates directly such as with geopandas and divide them by 100000?
The data is most likely in a specific cartographic projection. You don't just want to divide by 100k - the data will likely have nonlinear transformations which have a different effect on the position depending on the location. See the GeoPandas docs on working with projections.
If the CRS of the data is correctly encoded, you can re-project the dataframe into lat/lons (e.g. WGS84, which has the EPSG Code 4326) using geopandas.GeoDataFrame.to_crs, e.g.:
df_latlon = df.to_crs("epsg:4326")
I'm trying to project some point data in ArcMap.
I created the point data in excel, then summarised it in R and exported it as a csv. Then added it through 'add XY data' tool in ArcMap, and exported it from an events file to a shapefile.
The points appear in the correct place, relative to each other, but when tried to overlay on the world imagery basemap, it simply sits down near Antarctica.
I've checked and re-checked the projections of the imported layer and the dataframe (I'd like to work in GDA_1994_MGA_Zone_56), but have also tried changing to a geographic coordinate system, which doesn't change anything.
In fact, when I try to change the projection to a gcs, using the 'project - data management' tool, this error comes up.
invalid extent for output coordinate system
Failed to execute (Project).
I can't for the life of me figure out what is going wrong, there's probably a simple explanation, but I'm at my wits end!
It helps to have them stored originally as numeric data, but since it's actually coming in as a CSV, I think they are importing as numeric -- as you note, they are correct relative to each other.
If your points are showing up on a map but in the wrong location, then their initial projection definition is incorrect. Looking at the Add XY Data dialog, you need to specify an initial coordinate system. If nothing is specified, ArcMap assumes that it is latitude/longitude decimal degrees.
Once it's imported as the wrong coordinate system, attempts to reproject will just continue the But you do not necessarily need to delete the points and start to re-import from scratch. Try using the Define Projection tool (which modifies the metadata to the desired projection) instead of Project (which mathematically recalculates coordinates from current projection to a new projection).
I'm building a web application using Angular and mapbox-gl-js as map library.
The map tiles are projected using EPSG:4326 Plate-Carrée and the dataset is GeoJson (WGS84).
Since mapbox supports only web mercator projection, the data is not displayed in the correct position on map.
Any suggestions how to reproject the dataset so mapbox is able to display the data in the correct location?
Thanks.
If I understand your question correctly, you have some raster tiles which are "projected" (well, not really) in EPSG:4326, and you would like to display them in Mapbox-GL-JS. You also have another dataset, in GeoJSON, also provided in EPSG:4326, which you would like to overlay.
As you note, Mapbox-GL-JS only supports the EPSG:3857 (Web Mercator) projection. It projects datasets (but not raster tiles) from EPSG:4326 to EPSG:3857 in order to display them.
Your good options are:
Find a different basemap in Web Mercator
Find a web service which can reproject the tiles you have on the fly
Maybe there's some way where you could let Mapbox-GL-JS think it's displaying your tiles in EPSG:3857, then somehow adjust the coordinates of your overlaid dataset so that when they're reprojected from EPSG:4326 to EPSG:3857 they end up in the right positions. But I can't think how to achieve that.
You could try asking at gis.stackexchange.com.
these vector or raster files being classic files without geocoordinates. They are lat/long projection, I want to import them into QGIS, scale them up/down, place them to their right place, and they become reusable shp or raster geocoordinated layers.
Edit: I'am from the wikipedia Graphic Lab>Map workshop, we want to work more using GIS. We litteraly have hundreds maps to migrate to GIS technologies....
File:Chinese_plain_5c._BC-en.svg
File:Vignobles_basse_loire.svg
Partial Solution: load SVG into Inkscape, Save as DXF file, then you can load this into QGIS. This should at least get you most of the linework into QGIS.
However, it won't yet be properly georeferenced or styled, and different layers may be in different places because the SVG has some scaling and translating operators on parts of the map data that QGIS or Inkscape is ignoring. You'll probably need to work with a layer at a time. This probably isn't a problem since maybe you are only interested in the added data on the map, and not the base map (country outlines etc) since you will probably want to overlay your data onto standard map base layer (natural earth, OpenStreetMap tiles).
The only way I see to do the transformation at present is to work out the affine transformation parameters and use the QgsAffine plugin, but that does require you to work out the parameters beforehand by fitting known source coordinates to known target coordinates.
But to do hundreds? You might be better off writing some custom SVG parsing code for each one...
If you only want to display it in the correct place, scale and rotation treat it as an SVG icon.
1. create a point layer and put a single point at the georeferenced centre of the SVG you will load.
2. edit the symbology and load the SVG as an icon
3. set the size units to map units
4. supply the appropriate dimensions
5. rotate as necessary
The redraw is very slow and painful, but if you use Project>import/Export>Export map as image you can make a georeferenced raster.
I have a map, which is accurate (enough) to the real world map, as viewed in google maps.
I'm projecting the drawn image using the google.maps.ImageMapType class, thus the corners of the new, drawn, map are (sw)-90,-180 and (ne)90,180.
Now, knowing the coordinates of the south-west and north-east of the overlay in the real-world map, how do I convert GPS coordinates to this new scale?
OK, I added a marker in "Paris" to your jsFiddle, but the marker is clearly not where it should be.
I think you might be mixing google.maps.MapType object and google.maps.ImageMapType
google.maps.ImageMapType uses a google.maps.ImageMapTypeOptions object, which should have a getTileUrl() method, but instead you have a getTile() method which is breaking the projection.