Radiant Maps

Time to make your own beautiful maps: sign up to the mailing list to get a free .AI of map patterns. You'll also be notified when the 36 Square Inches of Map course is ready; it'll be the only end-to-end mapping course by a National Geographic-published cartographer.

Some fantastic cartographers:

Inspo:

The most beautiful maps are not behind us: they will be made today, by people who cared enough to make them. That sounds like you.

Great maps require no formal training or credentials. I learned by trial and so will you.

The best contemporary mapmakers taught themselves: Eleanor Lutz is a bio PhD and learned cartography on her own. Illustrator Mike Hall sketched maps in his notebook during shifts as a security guard, and now has an agent to handle his map deals. Alex McPhee studied geophysics, decided he needed to map his native Alberta, got some open source geo-software and made one of the best modern reference maps.

Mapmaking is like cooking: you pick what you like to eat, and improve by trial. You’ll over-salt a few dishes, ruin a few pans and come out a master. You just have to pick some Territory to commemorate.

Just draw one!

Cartography is making the infinite Territory legible to humans, which gives you many ways to get to the same point.

Find some Territory you care about: your yard, your street, your neighborhood, your town, your favorite patch of woods. Get a pencil and paper. Mark what you care about. You’re now a cartographer. If you stopped reading and drew a map on a paper towel, I’d be satisfied.

You can make a map with charcoals and an easel, a stick and some sand, a pencil and graph paper, a grid computing cluster, some hideously expensive software, some there’s-no-way-this-is-free software, a drone with a camera, a satellite with a radiometer, all ways to the same end. There’s no “correct” way to make a high-effort map.

If you still want to use the computer to make a nice-looking map, read on.

Disclosure

I learned cartography at close range from Daniel Huffman. If possible close this tab, get a desk next to a cartographer, pester with questions for 1+ year.

Every cartographer has their own idiosyncratic way of making maps; this is only what I can get my head around. If this doesn’t work for you please check out other ’tographer tutorials at the bottom, they might click better.

My entire toolkit:

QGIS
GDAL
MAPublisher
Carto
Mapbox Studio
Photoshop
Illustrator
InDesign
ImageMagick

⓿ Getting started

Here’s my map workflow, you’ll find yours soon enough:

Download geographic data: roads, rivers, lakes, hills, valleys, towns, lighthouses, etc.
Thin out, clean, futz with that data in free software programs QGIS and GDAL. Modern start-with-pile-of-data carto is about knowing what to leave off the map, so you’ll spend a lot of time sifting tangles of data for what you actually want to show.
Make the remaining data look nice in your graphic design software.

▼ There’s a big directory of links at the bottom of this page; if you need data, a tutorial, some inspo, check down there first. ▼

The Spatial Community – The upsetting part about learning cartography is the software, so you need computer pals to ask “why’s this broken?” The Spatial Community is a Slack channel where technical geographic information system (GIS) experts hang out, very friendly people who don’t mind questions from new mappers. Request an invite then join #cartography-designs, #gdal-ogr, #qgis, #data-sources and #newcomer-questions. Please ask questions when you get stuck! We’re here to help you!
Map School – glossary of map jargon.
Map Resources Page – links compiled by Robin Tolochko.
If you’re on twitter, search #cartography or #practicarto on twitter and start asking questions.

❶ Install some free software

Install click-around map software

QGIS – download QGIS, an open source geographic program. Lets you manipulate geodata with relative ease and export an SVG/PDF to cute up in Illustrator.

QGIS tutorial 1 – getting started with QGIS.

QGIS tutorial 2 – more QGIS tutorials.

Install type-around map software

You’re gonna install GDAL (Geospatial Data Abstraction Library), a command-line tool that lets you edit geodata without clicking around in QGIS. Learning to type your way through changing projections, cropping images, filtering data etc. will greatly speed up your mapping. It might feel awkward at first but it’s very useful; I hate coding and I still love command-line tools.

Using Windows?

Install miniconda, a “package manager.” Head here > download and run “Miniconda3 Windows 64-bit.”
After it’s done installing, find and run “Anaconda Prompt.”
Type “conda create --name mapenvironment”, without quotes. You can change “mapenvironment” to whatever you’d like; this is just creating an “environment” in which to place your software.
Type “conda activate mapenvironment” without the quotes.
Type “conda install gdal” without the quotes. You should see a bunch of text flow by.
To see if it worked, type “gdalinfo --version” without the quotes. You should see something like “GDAL 3.0.3, released 2020/01/08.”
When you want to use GDAL: run “Anaconda Prompt,” type “conda activate mapenvironment,” and now you can access tools like “gdalinfo” and “gdal_translate”

Using a Mac?

Install miniconda, a “package manager.” Head here > download and run “Miniconda3 MacOSX 64-bit pkg.”
After it’s done installing, open the Terminal app.
Type “conda create --name mapenvironment”, without quotes. You can change “mapenvironment” to whatever you’d like; this is just creating an “environment” in which to place software.
Type “conda activate mapenvironment” without the quotes.
Type “conda install gdal" without the quotes. You should see a bunch of text scroll by.
To see if it worked, type “gdalinfo --version” without the quotes. You should see something like “GDAL 3.0.3, released 2020/01/08.”
When you want to use GDAL: run Terminal, type “conda activate mapenvironment,” and now you can access tools like “gdalinfo” and “gdal_translate”

❷ Find some geographic data

You’ll encounter dozens of arcane file types but here are the main ones to look out for. All of these get worked over in QGIS and GDAL.

.SHP – Shapefile – stores vector shapes + associated data. Imagine you drew a hexagon and saved it as a vector file, but didn’t stop there: you also attached a table to that shape listing the hexagon’s name, when you drew it, its size in square meters, etc.
.GEOJSON – GeoJSON – also stores vector shapes + data tables.
.TIFF – GeoTIFF –stores raster data like satellite images, terrain data. Like a regular TIFF except it comes with georeferencing that tells QGIS where on earth to place it.
.GDB –GeoDatabase – Stores vector shapes + data too.

Where to get geographic data

Countries, Lakes, Rivers – Natural Earth supplies geographic boundaries and features; for a quick overview download the quick start kit, open the .QGS file in QGIS and see how it looks. Not great? Well, thats where you come in.
Roads – Natural Earth has roads, but not all of them. If you want to make a road map of your city, use OpenStreetMap data downloaded via the BBBike.org site. You zoom to where you want some data, draw a box, give it your email and it’ll send you a shapefile that you can view in QGIS.
Land cover (is that part of earth a forest? A town? A desert?) – Want to show what’s down there? If your map shows a big chunk of the world, 300-meter resolution Globcover 2009 data should work; find “A coloured version of the map in GeoTIFF format” on that page to get a file you can tune up in QGIS and Photoshop. For North America the Commission for Environmental Cooperation has a detailed 30-meter resolution dataset.
Terrain – the hills and valleys, a.k.a “shaded relief” or “hillshade.” This used to be drawn with graphite and airbrushes, now you turn satellite altimetry data into little computer-generated landscapes. You can get pre-generated terrain from Natural Earth: for zoomed-in maps, for zoomed-out maps, for just the U.S..

You can also roll your own terrain by grabbing elevation data for your map area from the Open Terrain Project and making your own (see step ❸).
Bathymetry (underwater terrain) – If your map includes undersea terrain, use the GMRT map tool to download data. Pick the rectangle tool at the top > drag a rectangle for your area > “create grid file”, file format: geotiff, mask: unmasked, grid resolution: maximum > download grid. You can process this into contours or a hillshade in step ❸.

Need something else? Time to start sifting the resources page down there ▼.

❸ Design your map

So now you have QGIS and GDAL installed, plus a folder full of geographic data. How do you turn that into a map? Someday you’ll get an end-to-end account (it’d take like 30,000 words and good screencasts) but for now you join the grand tradition of “follow a tutorial (check the resources pile at the bottom of this page ▼), ask your computer pals at The Spatial Community for help when you get stuck.”

For now, some general guidelines:

Reproject your data

Whatever coordinate system you project your data into, you gotta apply that same projection to all the other data in your map. It’s all gotta match!

Vector data is easy to reproject in QGIS. Raster stuff like imagery and elevation data is a bit more involved:

You can get into the weeds on picking a map projection, since you’re on the leaves of the old math problem of how to represent 3D stuff on a 2D plane. Luckily nobody’s navigating by your map, so just pick a projection that looks nice. Draw out a box enclosing your area of interest on projectionwizard.org and see if any of those look good. Or just default to Universal Transverse Mercator (UTM), nobody complains about that one.
Get a projection you like? Copy the PROJ string from projection wizard or look up your UTM zone’s EPSG code on EPSG.io
Feed that string or EPSG code into GDAL to change the projection of your data. You can also change projections in QGIS.

Make the terrain

My favorite part of maps is the topography. Make your own in QGIS or...
Try it in GDAL. The command line is a bit painful to learn at first but it will be much faster; making nice terrain requires a lot of experimentation, so you’ll need that speed.
If you want very nice terrain, make it in a 3D-modeling program (Blender). It’s free and not that hard if you follow the instructions carefully.
Eduard turns your DEMs into a looks-like-an-Imhof-graphite-relief (ML......wow)
PLEASE draw your own terrain: Sarah Bell wrote two tutorials on how to render shaded reliefs in pencil.

Design time

Once your data’s cleaned up and you’re left with what you want to show on your map, you export an SVG/PDF out of QGIS and make your vectors look nice in Illustrator or Inkscape. For the rasters (e.g. terrain and satellite imagery) export a TIFF from QGIS/GDAL and edit them in Photoshop or GIMP. There are tools that make this more convenient, like the wildly expensive MAPublisher plugin for Illustrator, but they’re not necessary.

Find some maps you like and see how close you can get; I think of ’tography as more craft than art, so you can get real far by copying the masters. Raid the inspo column down there ▼.

Now you’re in the art zone: compositing in Illustrator, labeling, futzing with colors, upsetting back-tracks to your original geodata, adding cute ephemera like north arrows and legends, illustrations. Make it look nice. Real nice. I can’t wait to see your map ♡.

❹ Good luck deciphering my notes

SIXTEEN SQUARE INCHES OF MAP: A GUIDE BY CARTOGRAPHER EVAN APPLEGATE

Introduction: why make maps, who this guide is for (designer or illustrator)
- READ ME FIRST
  - You comfortable fighting with the computer and typing into something called an "Integrated Development Environment?" You want to make beautiful maps? You tolerate lots of trials? You have any graphic design/illustration talent? What follows: download links and meta-download instructions plus idiosyncratic user notes on how n = 1 computer cartographer makes his 51st percentile maps.
- Computer setup
  - I use OSX, but this stuff works on windows. Past the initial setup you'll see OSX stuff only.
- Installing software
  - QGIS
  - Design software
    - Inkscape/Illustrator
      - All commands are in Illustrator but Inkscape works about the same way
    - GIMP/Photoshop
      - Same for Photoshop
  - Other sofware
    - Eduard
      - great for making hillshades + has a great 20m global DEM grabber
    - MAPublisher
      - costs 2800 but if you make a lot of maps professionally might be useful to you. wont be covered in this course
- Command-line stuff, why its good and fast (this ones a maybe)
  - GDAL/OGR: I install this via miniconda, much easier to keep it updated
    - Windows GDAL
    - Mac GDAL
1. Map data
  - 1.x Projections 101 (very basics (since i dont know much beyond that)
    - basic 3D > 2D geometry, sacrificing one: angles, distances, relative direction, size
      - https://ihatecoordinatesystems.com/
    - What you have to remember: you gotta get all your data into the same coordinate reference system (CRS). Rasters, vectors, all of it.
  - Scale
    - what are you trying to show?
    - Large-scale: zoomed-in, smaller denominator
    - Small-scale: zoomed-out, larger denominator
  - 1.x Gather raster data, diff between vector and raster
    - 1.x.x Elevation data: collected by lidar/radar, basic concept
      - USGS sources
      - OpenTopography
      - viewfinder panoramas
      - Eduard
      - Undersea data
        
        GEBCO
        
        NOAA topobathy
    - 1.x.x Land cover data (link to sources, general method)
      - GLOBCOVER
      - NLCD can load this into QGIS as a WMS
      - that chinese dataset?
    - 1.x.x Satellite data: radiometers that can make images that look like they were collected by your DSLR CCD
      - Bands, indices
      - ESA sentinel explorer
      - USGS methods
      - QGIS method
  - 1.3 Gather vector data
    - 1.x.x Toponymy (GNIS, gazzetteer, etc) Easiest way to get stuff for the US: https://edits.nationalmap.gov/apps/gaz-domestic/public/search/names , tick visible in current extent, query, click download>arrow
      
      theres an API but i dunno how to use it
      
      mapping outside the US but want to know what shits called? can spatial-query + export with https://geonames.nga.mil/geonames/GeographicNamesSearch/ , desig_cd is the type of feature, lookup table is
      
      layer > add layer > add delimited text layer > ... enxt to filename, pick your CSV > under geometry definitions, select point coordinates >leave geometry CRS at 4326
      
      too many? research tools > select by location > first box is the GNIS point layer youre gonna cut donw, second box is the bounding box you made in the beginning > run, it'll select all the points > right click the GNIS point layer > export > save features as > tick save only selected features > change CRS to EPsG:26943 > format: ESRI shapefile (this doesnt really matter, geojson is just kinda slow for me)
      
      it'll give you all these categoreis, you dont need all of these on one map generally; making a physical geog map? keep the channels, points, bays, islands etc, ditch civil, military, etc
      
      census you can safely ditch
      
      Area Bar Basin Bay Beach Canal Cape Census Channel Civil Cliff Crossing Falls Flat Gap Gut Island Lake Military Pillar Populated Place Range Reservoir Ridge Sea Spring Stream Summit Swamp Valley Woods
      
      vector > data management tools > split vector layer > input layer is your GNIS points layer > uniqe ID field pick "class" > output file type geojson > output directory, pick a folder to store them > run, bunch of layers appear
      
      pick one of your new layers > symbology > size down the points to 0.25 labels > single labels > value set to feature name rendering > text, 2 point type (this makes them easier to thin out in AI) > rendering > overlapping labels > allow overlaps without penalty

styles > copy style > all style categories no composition in qgis, you do all that in AI new print layout > icon with + and a roll of paper, drag rect to canvas to add > item properties to change scale, final size > save as SVG > export layers as SVG groups > export text as text objects open in AI, you gotta get the point markers in the same layer as the labels (they're separate in the SVG) in the AI you can replace the placeholder raw relief with your land cover + relief, i exported a PNG * Undersea toponymy https://www.charts.noaa.gov/InteractiveCatalog/nrnc.shtml for US undersea stuff https://www.ngdc.noaa.gov/gazetteer/ Undersea Features Gazzetter https://www.ngdc.noaa.gov/mgg/bathymetry/maps/directdownload.html for older NOAA undersea charts, useful for labeling * 1.x.x Hydro (NHD, natural earth, census, FWS, OSM) https://www.census.gov/geographies/mapping-files/time-series/geo/tiger-line-file.html water > linear hydrography > pick state > pick counties > add SHPs to QGIS across two counties: download both, drag SHPs into QGIS > vector > data mgmt tools > merge vector layes > pick both > have the new layer in layers patlette vector > geoprocessing tools > clip > pick your all_area_water layer in first dropdown, your original extent polygon in the second > run right slick > export > save features as... > pick name, geojson, make sure the projection is same as the project EPSG:26943 now you have all linear hydrography (creeks, rivers) on one layer clipped to your map extent repeat same steps for area hydrography (lakes, ponds) merge all rivers, lakes of same same: vector > geoprocessing tools > dissolve > pick input layer what you want to work on > dissolve field is FULLNAME, 322 features to 89. repeat same for area hydrography coastline: https://www.census.gov/geographies/mapping-files/time-series/geo/tiger-line-file.html download national file drag SHP into QGIS vector > geoprocessing tools > clip > pick your all_area_water layer in first dropdown, your original extent polygon in the second > run vector > geoprocessing tools > dissolve > pick input layer what you want to work on > dont pick dissolve fields, this merged everything into one line * 1.x.x Political bounds (natural earth) * 1.x.x Roads, rails (OSM, census) Roads: https://www.census.gov/geographies/mapping-files/time-series/geo/tiger-line-file.html web interface download all roads > pick state > pick county across two counties: download both, drag SHPs into QGIS > vector > data mgmt tools > merge vector layes > pick both > have the new layer in layers patlette vector > geoprocessing tools > clip > pick your all_roads layer in first dropdown, your original extent polygon in the second > run right click > save as geojson > right click > open attribute table > see RTTYP? here's the lookup C County I Interstate M Common Name O Other S State recognized U U.S. I, S, U are the larger roads, usually C is two-lane highways M, and Null (no RTTYP) are smaller vector > data management > split vector layer > unique ID field RTTYP > output file type geojson > pick an output directory > run now you have all roads on discrete layers * 1.3.4 Airport points, polys (OSM) * 1.3.5 Neighborhoods (gazzetteer? airbnb zillow?) * 1.3.6 Nature (OSM, parks, etc) * 1.3.7 Admin divs (US only) * 1.4 Processing your data * 1.4.1 Remember to reproject! * Raster in QGIS, GDAL * Vector in QGIS, GDAL * 1.4.2 Filtering Make a jig: QGIS create layer > new shapefile layer > pick a folder, name > geometry type point > it appears in the layer panel, > edit button > add point feature > click center of what you wanna map > click edit button agian to save > processing toolbox " rectangles, ovals, diamonds" > input layer = the shp you just made > shape "rectangle" > pick height, i did 50x50 mi TK > segments 4 > run > appears on layers panel > right click export > save features as... > pick a file name> make sure CRS matches bottom right of your QGIS window, the project CRS right click > layer properties > information > Information from provider > Extent > copy those four values, chop off last digits 1846577.07,502154.17 : 1904513.46,560090.55 rearrange to ??? head to https://coast.noaa.gov/dataviewer/#/lidar/search/-13618597.819984838,4374679.235509932,-13544447.333741099,4451287.649944467 to get an MTY topobathy DEM pick NOAA NCEI Continuously Updated Digital Elevation Model (CUDEM) - Ninth Arc-Second Resolution Bathymetric-Topographic Tiles https://noaa-nos-coastal-lidar-pds.s3.amazonaws.com/dem/NCEI_ninth_Topobathy_2014_8483/index.html download tile index Tile Index: tileindex_NCEI_ninth_Topobathy_2014.zip open > shp into QGIS window > select features by area (make sure the new layer is selected on the left) > drag box around the tiles tht overlap your square > right click on tiles layer > open attribute table > bottom left "show selected features" right click tile index layer > export features > pick CSV > tick "save only selected featueres" > open CSV > copy list of URLS > paste in a plain text file > download each one by one or wget -i your_file_list.txt drag all .TIFs into their own dir `gdalbuildvrt DEM_huge.vrt *.tif` drag this into window > dlb clickl to open layer properties > see pixel size = 3, -3, so its 3m pixels, ~10ft res topobathy DEM. thats a little big for the map we're making resize, resample, crop and reproject that big DEM: `gdalwarp -co "COMPRESS=LZW" -tr 10 10 -r cubic -t_srs EPSG:26943 -cutline 36_mile_square.shp -crop_to_cutline out.vrt cropped_10m.tif

drag this into window > dlb clickl to open layer properties > see pixel size = 3, -3, so its 3m pixels, ~10ft res topobathy DEM. thats a little big for the map we're making if oyu're dumping these into EDUARD you can use whatever size you want since you can interactively generalize processing > terrain shading to make your outputs, test a bunch with ambient occlusion hillshade shadow depth texture shading when its time to go to photoshop we gotta take these from 32 bit to 8 bit to make them easier to manp in PS; theyre not RAWS you dont need all the range right click > export > save as > tick "rendered image" Or use bbox.io? * Raster * Cropping * Bit depth and re-scaling DNs * Relief design * Eduard * QGIS * Terrain Shading Plugin * Hillshade * TPI * Texture shade * Shadow depth * GDAL * Different flags * Blender * Link out, I hate this thing * Vector * Spatial queries * Attribute-table queries * Contours check DEM units -i is the interval; so -i 20 on a meter-unit DEM means your contours will have 20m intervals gdal_contour -a ELEV -i 40 -f "ESRI Shapefile" in1.tif "out_1" I recommend using this script by Henrik https://hkartor.se/anteckningar/contour_lines_script.html I modded it to change the -ot to Float32 so I can use it for underwater bathy if you want contour polygons stedda lines use `-p -amin "min_elev" -amax "max_elev"` stedda `-a ELEV` to label contours https://opensourceoptions.com/how-to-create-contour-lines-and-labels-with-qgis/ only thing id add: placement > general settings > mode: curved note: this breaks up your text paths. QGIS SVG export doesnt understand type on a path

1. Map Design
  - Naked theft: NPS etc
    - Choose a state, choose a park, get an Illustrator version of that map: https://www.nps.gov/carto/app/#!/parks
    - Metadata: https://www.nps.gov/carto/app/#!/maps/production-tips
    - .AIT templates towards the top, broken links but if you paste them into the archive.org waybackmachine they download fine: https://www.nps.gov/carto/app/#!/maps/starter-maps , e.g. https://web.archive.org/web/20170605123357/https://www.nps.gov/hfc/carto/starter-maps/a_starter.zip
  - 2.1 Setup
    - Layers
    - Do as much as possible in QGIS to reduce one-way trips
      - create extent line that's always on your exports
    - Gestalt first!
      - lookit stuff liek this first
  - X.1 Vector
    - X.1 Color
      - Steal from better maps
    - X.1 Type Arranging type is the most labor-intensive part of cartography. A good map has no collisions, no ambiguous placement, not one careless label. It's the brown M&M test of mapmaking: if the maker didn't look over every square inch, it'll show in the labels.
      - Examples Mike Hall, Daniel Huffman, Michelle Snyder, Marty Schnure, Jeff Clark, Dave Imus, Carl Churchill, Alex McPhee
      - Halos Labeling tip from Andy Woodruff: instead of using outer glow/drop shadow/blurred strokes to set labels off from the background, use a blurred version of your raster underlayer that shines through 0% opacity strokes to a blurred version of your raster.

1) "Burn" your rivers, roads into your raster, You can do this in Illustrator by exporting a no-labels PNG of the whole map, artboard = full extent. 2) Create a blurred version of this raster: I like turning it into a Photoshop smart object before gaussian-blurring so I can adjust it later. 3) Layer order top-to-bottom: all map stuff (including your raster underlayer) in one supra-layer > blurred raster 4) Check knockout group on your supra-layer 5) For your label layer: add a 0% opacity blurred stroke, drag it under "contents" to ensure your original type colors are preserved 6) Enjoy your sweet movable type halos * Knockouts https://somethingaboutmaps.wordpress.com/2021/05/19/blurring-backgrounds-to-improve-text-legibility/ https://somethingaboutmaps.wordpress.com/2015/05/08/type-knockouts-in-illustrator/ https://somethingaboutmaps.wordpress.com/2016/12/05/even-fancier-type-knockouts-in-illustrator/ https://somethingaboutmaps.wordpress.com/2018/10/28/smart-type-halos-in-photoshop-and-illustrator/ * How to label... * Rivers, Ranges, Areas type on a path is annoying but no way to improve at this without getting your reps in. Use direct select tool (A, looks like a white arrow) to drag the handles from either end, if your text flips upside down drag ⟘ perpendicular to your line, if you can't see anything but a red + you gotta drag the handles from either end. Add points to the line with pen tool add anchor point, drag out sides, then use direct select to select new anchor point, drag bezier handles to get the text lookin right. You gotta track this out; a long mountain range or river gets type that looks like t h i s and a short one gets text that looks like this. * Points top right? * Road shields steal from noun project * Soundings itals * Contours dunno how to do this automated * X.1 Icons * Symbol palette Use Nate Kelso's replace-with-symbol to get QGIS's lil point symbols into symbols; create symbol you like, select it + the points you want to turn into that symbol, run this https://kelsocartography.com/scripts/scripts/nvkelso/findAndReplaceGraphic_centered_v2.jsx * X.1 Lines * Casing roads * Dotted, dashed lines * X.1 Polygons/shapes/fills * Glows Polygons with interior faded glows: black fill, feather effect, opacity 0% above ANOTHER fill with the color you like. might need knockout group enabled too * Coast lines (adjustment panel) Making vintage-style coast lines in Illustrator: to your coastline path add a series of strokes in the appearance panel, each gets an offset path effect. Increase the offset path distance between each stroke so they spread out as they get further from the coast. This is a last step as it makes Illustrator very slow; I save the appearance as a graphic style and apply it when I'm ready to export. * Generalization Cut down voids/gaps in your parks layers manually, makes a cleaner output. mapshaper.org is good for this too if you jstu want fewer vertices * Metadata in QGIS print composer on the left is "add scale bar," drag a box and it'll appear, i change it to miles and add some segments. you'll use this as a template to draw over in illustrator * X.1 Raster * Blend modes, opacity * Adjustment layers * Land cover NLCD WMTS layer > cropped TIF > PS > re-paste into discrete classes > change colors, add patterns from my old LC.psb > steal colors from henrik, churchill, old USGS > use end-at-land topobathy DEM to mask land and add topobathy gradient NLCD to vector: If you want e.g. wetlands, use the QGIS raster calculator to get classes 91 and 92, "woody wetlands" and "emergent herbaceous wetlands," into their own TIFF. raster calc > ("NLCD land cover@1" = 91) OR ("NLCD land cover@2" = 92) Use QGIS polygonize to turn those classes into polygons Delete polygons where DN = 0, those are the ones we dont want Exported to SVG Place polygons in Illustrator, offset path to expand them, make everything into a compound path, use that compound path as a mask over a pattern of marsh symbols. I never have any luck at making repeating patterns so I just copy/pasted a bunch of little marsh marks. Can also use the built-in pattern swatches; Swatches > Open Swatch Library> Patterns > Basic Graphics > Basic Graphics_Textures, I like "USGS 17 Dry Sandy Lake." otehrs are nice too * Color schemes * Ice * Patterns * Elevation * Relief layer order * The futz... * Imagery drape

Appendix
- Learn from these tutorials

1.1 Basic Raster Processing 1.1.1 Merging Tiles Exercise 1 1.1.2 Converting Formats 1.1.3 Compressing Output 1.1.4 Setting NoData Values 1.1.5 Writing Cloud-Optimized GeoTIFF (COG) 1.2 Processing Elevation Data 1.2.1 Creating Hillshade 1.2.2 Creating Color Relief Exercise 2 1.3 Processing Aerial Imagery 1.3.1 Create a preview image from source tiles 1.3.2 Create a Tile Index 1.3.3 Mosaic and clip to AOI 1.3.5 Creating Overviews 1.4 Processing Satellite Imagery 1.4.1 Merging individual bands into RGB composite 1.4.2 Apply Histogram Stretch and Color Correction 1.4.3 Raster Algebra Exercise 3 1.4.4 Pan Sharpening 1.5 Processing WMS Layers 1.5.1 Listing WMS Layers 1.5.2 Creating a Service Description File 1.5.3 Downloading WMS Layers Exercise 4 1.6 Georeferencing 1.6.1 Georeferencing Images with Bounding Box Coordinates 1.6.2 Georeferencing with GCPs Exercise 5 Assignment

OGR Tools 2.1 ETL Basics 2.1.1 Read a CSV data source 2.1.2 Convert it to point data layer 2.1.3 Assign it a CRS 2.1.4 Extract a subset 2.1.5 Change the data type of a column 2.1.6 Rename the layer in GeoPackage. Exercise 6 2.2 Merging Vector Files Exercise 7 2.3 Geoprocessing and Spatial Queries 2.3.1 Reprojecting Vector Layers 2.3.2 Creating Buffers 2.3.3 Performing Spatial Queries 2.3.4 Data Cleaning
Running commands in batch
Automating and Scheduling GDAL/OGR Jobs Tips for Improving Performance Configuration Options Multithreading Supplement Check Supported Formats and Capabilities Extracting Image Metadata and Statistics Validating COGs Creating Contours Creating Colorized Imagery Creating Colorized Hillshade Removing JPEG Compression Artifacts Splitting a Mosaic into Tiles Extracting Projection Information from a Raster Merging Files with Different Resolutions Calculate Pixel-Wise Statistics over Multiple Rasters Extracting Values from a Raster Raster to Vector Conversion Viewshed Analysis Working with KML Files Exporting Data to KML files Converting KML Files to Other Formats KML vs. LIBKML Drivers Group Statistics Using Virtual Layers Read Geonames Files Applying Filters Merging Files Resources Data Credits License

Step zero

💬 Join The Spatial Community Slack and get answers to your geo-questions.

QGIS: open source GIS software for Windows, OSX, Linux

GDAL/OGR: open source command-line GIS tools

Natural Earth: public domain data source for borders, countries, cities, natural features, and more

You into CLI + Python, R, C++? Here's every geodata processing tool

Resource collections
By Maptime
By Robin Tolochko
By dragons8mycat
By RT Wilson

Inspiration

Library of Congress (change search dropdown to maps)
Historical NOAA nautical charts
U.S. National Arch ives
American Geographical Society Collection
National Library of Scotland
British Library
Boston Library
David Rumsey Map Collection
USGS historical topographic maps
Public domain shaded reliefs
New York Public Library map collection
Perry-Castañeda Library Map Collection
Osher Map Library

Some fantastic cartographers:

Data sources: land cover, elevation

Land cover
30m U.S. land cover (NLCD)
30m U.S. croplands
30m North American land cover
30m global land cover
100m global land cover
300m global land cover

Elevation
100m CONUS shaded relief + land cover
30-90m elevation data
Elevation data finder: openterrain
Elevation data finder: opendem
Elevation data finder: opentopography
Elevation data finder: imagico
Bathymetry finder
Antarctic bathymetry
U.S. wildfire perimeters

Data sources: satellite imagery

Remote sensing basics
250m Blue Marble
Sentinel-2 + Landsat 8 browser: Sentinel Playground
Sentinel-2 + Landsat 8 browser: EO Browser
MODIS/VIIRS browser
GOES-16 browser
Sentinel-1 browser
Download by lat/long from all of NOAA’s sensors
High-resolution imagery browser (expensive)
Tim Wallace’s satellite imagery resources
How to find the most recent imagery
Charlie Loyd’s imagery compendium
How's your javascript? Google Earth Engine has all sat data and you can do anything you want to ten billion pixels at once.

Data sources: street-level

Download OpenStreetMap raster tiles
Download OSM data by bounding box
Download Overture Maps Foundation data by bounding box (I wrote this in an hour using GPT4)
Download OSM data by city
Download OSM data by country
Building footprints for the U.S., Canada, Tanzania, Uganda
Global city boundaries

Demographics
U.S. Census geodata