Return to Answer

Something most likely went wrong with your projection settings along the way, but it's a bit hard to figure out where exactly the error lies, so it might be easier to walk through the process from the start.

1. When you load the shapefile, you'll notice that you're asked to pick the coordinate reference system (CRS) your shapefile is in. Usually, this dialog won't pop up, because the correct coordinate system is stored in one of the files that make up the shapefile: the one ending in .prj. That one's missing with your dataset, however, hence the dialog. So we need to find out what CRS your data is in. Without this information, any projections we apply later on will be meaningless.

2. Googling for rs14fe02 (the base name of your shapefile) points to the NOAA metadata entry. The section "Spatial reference information" tells us that the data is in geographic coordinates (i.e. it uses latitude and longitude, not distance units), and that it uses the North American Datum of 1927 with the Clarke 1866 ellipsoid. Scrolling through the list of CRS's in the QGIS dialog, we find an entry called NAD27 in the "Geographic Coordinate Systems" section. The proj string is +proj=longlat +ellps=clrk66 +towgs84=-10,158,187,0,0,0,0, so that matches the metadata.

3. After choosing the NAD27 CRS and clicking OK, the shapefile is opened:

   

   I've added a graticule of longitude and latitude: As you can see, they're perfectly square, which is what you'd expect from a geographic coordinate system displayed in QGIS, since it just uses an equirectangular projection to display the data. If you measure a distance using the line measure tool, it's in degrees, which again, is what you'd expect, but not what you want for buffering using linear units.

4. So we need to project the data. For buffering, we should use an equal area projection to avoid introducing errors into the analysis. A common one is the "US National Atlas Equal Area" (EPSG:2163), which is a Lambert azimuthal equal area projection. You can see what your data looks like in that projection by clicking on the globe in the lower right corner, activating "on the fly projection", and choosing EPSG:2163 from the list:

   

5. You might think that you could just create the buffer now, but you'd be disappointed if you did: The geoprocessing tools don't care what projection is selected for the project, they care what projection the data is in. We haven't changed that: The shapefile is still in NAD27 geographic coordinates, so running a buffer with a buffer distance of 3000 "units" would be interpreted as 3000°, not 3000 metres. Instead, we explicitly have to project the data first: Select the layer, and click "Layer / Save as...", choose a file name, and choose "Selected CRS" and "EPSG:2163". Tick "Add saved file to map" and click OK. After the file is saved, QGIS will load and display it.

6. Now you can run the buffer with the proper units: Open the "Vector / Geoprocessing / Buffer" dialog, choose the newly created layer, and enter the buffer distance in metres. Once it's done, the layer will be added to the canvas.

   Here's a section from the north eastern US with a 5 km buffer:

   

Something most likely went wrong with your projection settings along the way, but it's a bit hard to figure out where exactly the error lies, so it might be easier to walk through the process from the start.

1. When you load the shapefile, you'll notice that you're asked to pick the coordinate reference system (CRS) your shapefile is in. Usually, this dialog won't pop up, because the correct coordinate system is stored in one of the files that make up the shapefile: the one ending in .prj. That one's missing with your dataset, however, hence the dialog. So we need to find out what CRS your data is in. Without this information, any projections we apply later on will be meaningless.

2. Googling for rs14fe02 (the base name of your shapefile) points to the NOAA metadata entry. The section "Spatial reference information" tells us that the data is in geographic coordinates (i.e. it uses latitude and longitude, not distance units), and that it uses the North American Datum of 1927 with the Clarke 1866 ellipsoid. Scrolling through the list of CRS's in the QGIS dialog, we find an entry called NAD27 in the "Geographic Coordinate Systems" section. The proj string is +proj=longlat +ellps=clrk66 +towgs84=-10,158,187,0,0,0,0, so that matches the metadata.

3. After choosing the NAD27 CRS and clicking OK, the shapefile is opened:

   

   I've added a graticule of longitude and latitude: As you can see, they're perfectly square, which is what you'd expect from a geographic coordinate system displayed in QGIS, since it just uses an equirectangular projection to display the data. If you measure a distance using the line measure tool, it's in degrees, which again, is what you'd expect, but not what you want for buffering using linear units.

4. So we need to project the data. For buffering, we should use an equal area projection to avoid introducing errors into the analysis. A common one is the "US National Atlas Equal Area" (EPSG:2163), which is a Lambert azimuthal equal area projection. You can see what your data looks like in that projection by clicking on the globe in the lower right corner, activating "on the fly projection", and choosing EPSG:2163 from the list:

   

5. You might think that you could just create the buffer now, but you'd be disappointed if you did: The geoprocessing tools don't care what projection is selected for the project, they care what projection the data is in. We haven't changed that: The shapefile is still in NAD27 geographic coordinates, so running a buffer with a buffer distance of 3000 "units" would be interpreted as 3000°, not 3000 metres. Instead, we explicitly have to project the data first: Select the layer, and click "Layer / Save as...", choose a file name, and choose "Selected CRS" and "EPSG:2163". Tick "Add saved file to map" and click OK. After the file is saved, QGIS will load and display it.

6. Now you can run the buffer with the proper units: Open the "Vector / Geoprocessing / Buffer" dialog, choose the newly created layer, and enter the buffer distance in metres. Once it's done, the layer will be added to the canvas.

   Here's a section from the north eastern US with a 5 km buffer:

   

Something most likely went wrong with your projection settings along the way, but it's a bit hard to figure out where exactly the error lies, so it might be easier to walk through the process from the start.

1. When you load the shapefile, you'll notice that you're asked to pick the coordinate reference system (CRS) your shapefile is in. Usually, this dialog won't pop up, because the correct coordinate system is stored in one of the files that make up the shapefile: the one ending in .prj. That one's missing with your dataset, however, hence the dialog. So we need to find out what CRS your data is in. Without this information, any projections we apply later on will be meaningless.

2. Googling for rs14fe02 (the base name of your shapefile) points to the NOAA metadata entry. The section "Spatial reference information" tells us that the data is in geographic coordinates (i.e. it uses latitude and longitude, not distance units), and that it uses the North American Datum of 1927 with the Clarke 1866 ellipsoid. Scrolling through the list of CRS's, we find an entry called NAD27 in the "Geographic Coordinate Systems" section. The proj string is +proj=longlat +ellps=clrk66 +towgs84=-10,158,187,0,0,0,0, so that matches the metadata.

3. After choosing the NAD27 CRS and clicking OK, the shapefile is opened:

   

   I've added a graticule of longitude and latitude: As you can see, they're perfectly square, which is what you'd expect from a geographic coordinate system displayed in QGIS, since it just uses an equirectangular projection to display the data. If you measure a distance using the line measure tool, it's in degrees, which again, is what you'd expect, but not what you want for buffering using linear units.

4. So we need to project the data. For buffering, we should use an equal area projection to avoid introducing errors into the analysis. A common one is the "US National Atlas Equal Area" (EPSG:2163), which is a Lambert azimuthal equal area projection. You can see what your data looks like in that projection by clicking on the globe in the lower right corner, activating "on the fly projection", and choosing EPSG:2163 from the list:

   

5. You might think that you could just create the buffer now, but you'd be disappointed if you did: The geoprocessing tools don't care what projection is selected for the project, they care what projection the data is in. We haven't changed that: The shapefile is still in NAD27 geographic coordinates, so running a buffer with a buffer distance of 3000 "units" would be interpreted as 3000°, not 3000 metres. Instead, we explicitly have to project the data first: Select the layer, and click "Layer / Save as...", choose a file name, and choose "Selected CRS" and "EPSG:2163". Tick "Add saved file to map" and click OK. After the file is saved, QGIS will load and display it.

6. Now you can run the buffer with the proper units: Open the "Vector / Geoprocessing / Buffer" dialog, choose the newly created layer, and enter the buffer distance in metres. Once it's done, the layer will be added to the canvas.

   Here's a section from the north eastern US with a 5 km buffer:

   

Something most likely went wrong with your projection settings along the way, but it's a bit hard to figure out where exactly the error lies, so it might be easier to walk through the process from the start.

1. When you load the shapefile, you'll notice that you're asked to pick the coordinate reference system (CRS) your shapefile is in. Usually, this dialog won't pop up, because the correct coordinate system is stored in one of the files that make up the shapefile: the one ending in .prj. That one's missing with your dataset, however, hence the dialog. So we need to find out what CRS your data is in. Without this information, any projections we apply later on will be meaningless.

2. Googling for rs14fe02 (the base name of your shapefile) points to the NOAA metadata entry. The section "Spatial reference information" tells us that the data is in geographic coordinates (i.e. it uses latitude and longitude, not distance units), and that it uses the North American Datum of 1927 with the Clarke 1866 ellipsoid. Scrolling through the list of CRS's in the QGIS dialog, we find an entry called NAD27 in the "Geographic Coordinate Systems" section. The proj string is +proj=longlat +ellps=clrk66 +towgs84=-10,158,187,0,0,0,0, so that matches the metadata.

3. After choosing the NAD27 CRS and clicking OK, the shapefile is opened:

   

   I've added a graticule of longitude and latitude: As you can see, they're perfectly square, which is what you'd expect from a geographic coordinate system displayed in QGIS, since it just uses an equirectangular projection to display the data. If you measure a distance using the line measure tool, it's in degrees, which again, is what you'd expect, but not what you want for buffering using linear units.

4. So we need to project the data. For buffering, we should use an equal area projection to avoid introducing errors into the analysis. A common one is the "US National Atlas Equal Area" (EPSG:2163), which is a Lambert azimuthal equal area projection. You can see what your data looks like in that projection by clicking on the globe in the lower right corner, activating "on the fly projection", and choosing EPSG:2163 from the list:

   

5. You might think that you could just create the buffer now, but you'd be disappointed if you did: The geoprocessing tools don't care what projection is selected for the project, they care what projection the data is in. We haven't changed that: The shapefile is still in NAD27 geographic coordinates, so running a buffer with a buffer distance of 3000 "units" would be interpreted as 3000°, not 3000 metres. Instead, we explicitly have to project the data first: Select the layer, and click "Layer / Save as...", choose a file name, and choose "Selected CRS" and "EPSG:2163". Tick "Add saved file to map" and click OK. After the file is saved, QGIS will load and display it.

6. Now you can run the buffer with the proper units: Open the "Vector / Geoprocessing / Buffer" dialog, choose the newly created layer, and enter the buffer distance in metres. Once it's done, the layer will be added to the canvas.

   Here's a section from the north eastern US with a 5 km buffer:

   

Something most likely went wrong with your projection settings along the way, but it's a bit hard to figure out where exactly the error lies, so it might be easier to walk through the process from the start.

1. When you load the shapefile, you'll notice that you're asked to pick the coordinate reference system (CRS) your shapefile is in. Usually, this dialog won't pop up, because the correct coordinate system is stored in one of the files that make up the shapefile: the one ending in .prj. That one's missing with your dataset, however, hence the dialog. So we need to find out what CRS your data is in. Without this information, any projections we apply later on will be meaningless.

2. Googling for rs14fe02 (the base name of your shapefile) points to the NOAA metadata entry. The section "Spatial reference information" tells us that the data is in geographic coordinates (i.e. it uses latitude and longitude, not distance units), and that it uses the North American Datum of 1927 with the Clarke 1866 ellipsoid. Scrolling through the list of CRS's, we find an entry called NAD27 in the "Geographic Coordinate Systems" section. The proj string is +proj=longlat +ellps=clrk66 +towgs84=-10,158,187,0,0,0,0, so that matches the metadata.

3. After choosing the NAD27 CRS and clicking OK, the shapefile is opened:

   

   I've added a graticule of longitude and latitude: As you can see, they're perfectly square, which is what you'd expect from a geographic coordinate system displayed in QGIS, since it just uses an equirectangular projection to display the data. If you measure a distance using the line measure tool, it's in degrees, which again, is what you'd expect, but not what you want for buffering using linear units.

4. So we need to project the data. For buffering, we should use an equal area projection to avoid introducing errors into the analysis. A common one is the "US National Atlas Equal Area" (EPSG:2163), which is a Lambert azimuthal equal area projection. You can see what your data looks like in that projection by clicking on the globe in the lower right corner, activating "on the fly projection", and choosing EPSG:2163 from the list:

   

5. You might think that you could just create the buffer now, but you'd be disappointed if you did: The geoprocessing tools don't care what projection is selected for the project, they care what projection the data is in. We haven't changed that: The shapefile is still in NAD27 geographic coordinates, so running a buffer with a buffer distance of 3000 "units" would be interpreted as 3000°, not 3000 metres. Instead, we explicitly have to project the data first: Select the layer, and click "Layer / Save as...", choose a file name, and choose "Selected CRS" and "EPSG:2163". Tick "Add saved file to map" and click OK. After the file is saved, QGIS will load and display it.

6. Now you can run the buffer with the proper units: Open the "Vector / Geoprocessing / Buffer" dialog, choose the newly created layer, and enter the buffer distance in metres. Once it's done, the layer will be added to the canvas.

Something most likely went wrong with your projection settings along the way, but it's a bit hard to figure out where exactly the error lies, so it might be easier to walk through the process from the start.

1. When you load the shapefile, you'll notice that you're asked to pick the coordinate reference system (CRS) your shapefile is in. Usually, this dialog won't pop up, because the correct coordinate system is stored in one of the files that make up the shapefile: the one ending in .prj. That one's missing with your dataset, however, hence the dialog. So we need to find out what CRS your data is in. Without this information, any projections we apply later on will be meaningless.

2. Googling for rs14fe02 (the base name of your shapefile) points to the NOAA metadata entry. The section "Spatial reference information" tells us that the data is in geographic coordinates (i.e. it uses latitude and longitude, not distance units), and that it uses the North American Datum of 1927 with the Clarke 1866 ellipsoid. Scrolling through the list of CRS's, we find an entry called NAD27 in the "Geographic Coordinate Systems" section. The proj string is +proj=longlat +ellps=clrk66 +towgs84=-10,158,187,0,0,0,0, so that matches the metadata.

3. After choosing the NAD27 CRS and clicking OK, the shapefile is opened:

   

   I've added a graticule of longitude and latitude: As you can see, they're perfectly square, which is what you'd expect from a geographic coordinate system displayed in QGIS, since it just uses an equirectangular projection to display the data. If you measure a distance using the line measure tool, it's in degrees, which again, is what you'd expect, but not what you want for buffering using linear units.

4. So we need to project the data. For buffering, we should use an equal area projection to avoid introducing errors into the analysis. A common one is the "US National Atlas Equal Area" (EPSG:2163), which is a Lambert azimuthal equal area projection. You can see what your data looks like in that projection by clicking on the globe in the lower right corner, activating "on the fly projection", and choosing EPSG:2163 from the list:

   

5. You might think that you could just create the buffer now, but you'd be disappointed if you did: The geoprocessing tools don't care what projection is selected for the project, they care what projection the data is in. We haven't changed that: The shapefile is still in NAD27 geographic coordinates, so running a buffer with a buffer distance of 3000 "units" would be interpreted as 3000°, not 3000 metres. Instead, we explicitly have to project the data first: Select the layer, and click "Layer / Save as...", choose a file name, and choose "Selected CRS" and "EPSG:2163". Tick "Add saved file to map" and click OK. After the file is saved, QGIS will load and display it.

6. Now you can run the buffer with the proper units: Open the "Vector / Geoprocessing / Buffer" dialog, choose the newly created layer, and enter the buffer distance in metres. Once it's done, the layer will be added to the canvas.

   Here's a section from the north eastern US with a 5 km buffer: