Category Archives: cartography

Waggawaggawaggawagga animated ducks in QGIS

I used ne_110m_admin_0_countries.

Rendering updates for the layer at 0.1 seconds.

Geometry Generator for a Point for the marker location via line_interpolate_point:

with_variable(
  'biggest_geom',
  geometry_n(order_parts($geometry, 'area($geometry)', ascending:=False), 1),
  line_interpolate_point(
    boundary(@biggest_geom), 
    perimeter(@biggest_geom)*(round(epoch(now())/100)%100/100)
  )
)

Raster Image Marker with https://opengameart.org/content/character-spritesheet-duck, vertical anchor at bottom, sprite choice between walking and running (doesn’t actually work) plus the frame via

with_variable(
  'biggest_geom',
  geometry_n(order_parts($geometry, 'area($geometry)', ascending:=False), 1),
  '/your/path/Duck/Sprites/Walking-Running/'
  || if(perimeter(@biggest_geom) < 10, 'Walking', 'Running')
  || ' 00'
  || to_string(round(epoch(now())/200)%2+1)
  || '.png'
)

Rotation did not work, I tried line_interpolate_angle:

with_variable(
  'biggest_geom',
  geometry_n(order_parts($geometry, 'area($geometry)', ascending:=False), 1),
  line_interpolate_angle(
    boundary(@biggest_geom), 
    perimeter(@biggest_geom)*(round(epoch(now())/100)%100/100)
  )
)

Steps via two more Geometry Generators, both for Lines using line_substring and some nice style (inspired by the wonderful built-in cat trail preset):

with_variable(
	'biggest_geom',
	geometry_n(order_parts($geometry, 'area($geometry)', ascending:=False), 1),
  	line_substring(
	  boundary(@biggest_geom), 
	  0,
	  perimeter(@biggest_geom)*(round(epoch(now())/100)%100/100)
	)
)

Could be improved if (for example) Raster Image Marker would support:

  • Choice of resampling algorithm
  • Flipping
  • Rotation would work, no idea what’s wrong with my expression, it works with random values, so …
  • Whatever is broken with the choice between ‘Walking’ and ‘Running’ in the file path expression

The effect of lossy LERC compression, visually “explained”

LERC is a kick-ass approach to 2D raster data compression, supported in GDAL since version 3.3. You can use it for lossless compression but it is also able to throw away some bits of information for smaller data sizes. You tell it which level of Z error is acceptable for your values and it will use that freedom to change the values of neighboring cells to do its magic. Z means the “data” axis here, of a single-band in a 2D raster, X and Y are the coordinates, or rather the locations of the data values in the raster, which are obviously not changed.

I thought it would be nice to show what it actually results in, you can read up on the details elsewhere if you want. Look at the images in full resolution please.

I used a global SRTM DEM with a Z value in full meters (no floating point values but integers) and applied LERC on it in three ways: Lossless, with a maximum Z error of 1 meter and a maximum Z error of 10 meters. Zstandard compression was always used.

The original GeoTIFF file was already very well compressed with Zstandard level 15 and a horizontal predictor; at ~1296000 x ~417600 pixels it has a size of 86 gigabytes including overviews.

  • Original (ZSTD level 15): 86 GB
  • LERC_ZSTD (lossless): 105G
  • LERC_ZSTD (maximum Z error of 1): 81G
  • LERC_ZSTD (maximum Z error of 10): 21G

Cool, so if we don’t care about an error of 10 meters, we can have a global DEM (well, as global as SRTM is with its 60° cut-off) at ~30 meters pixel resolution in 21 gigabytes. But what does that actually look like then and how will this error appear? Well, check it out:

Here are some samples visualised with a greyscale color ramp (locally adjusted, so the lowest value in the image is black, the highest value is white). They are shown at a 1:1 resolution, one pixel in the image (if you look at it at 100%) is one cell of the DEM data. The left image is lossless, the middle one was allowed an Z error of 1 meter, the right one 10 meters.

Mountainous, here the values range from 0 meters to about 2000 meters:

You can hardly see a difference, at least visually.

“Mediumish”, values between ~100 and ~500 meters:

At the 10 meter error level you can see a significant terracing effect.

Plains, values all around 100 meters:

You can see some structures collapsing into flat areas in the 1 meter version and oh wow that 10 meters version looks like upscaled pixels.

Time to zoom in! I picked a less flat area again because it makes it easier to understand. Here the values are between ~100 and ~300 meters:

So what do we see here? Neighboring cells with the same values compress better so LERC is shifting the values around (within the allowed error), creating terraces of same-valued cells. If you look closely you can see that there is also a visible pattern of squarish structures. Those are the blocks or windows in which LERC looks at the data and does its adjustments, in this case they were 8×8 pixels wide. Note: What LERC does exactly is a bit more complex than “try to make neighboring values the same”, it actually looks at the bits required to store the values within a block and optimized that within the error tolerance.

And now you know what LERC can do, if you give it a error level to play with.

For reference, here is that same-ish area with the error tolerance at 1 meter:

You have to zoom in quite a bit more to be able to see the effects here due to the nature of the data in this extent in combination with the particular error tolerance:

The larger the zonal differences of your Z values are to each other and in relation to the error tolerance, the less distinguished will this effect be. If there are steps of 100 meters between neighboring pixels, an extra error of 10 meters won’t do much of a difference. But in more flat areas it will have significant “terracing” effects as you could see above. This is similar to “banding” effects in images where there is little variation in color, e. g. a blue sky or an artificial color gradient, and you look at it in a setup that has a color bit depth on a resolution your human eyes can distinguish.

So if you want to use LERC with a lossy approach, think hard about what is going to happen with your data later. What kind of analysis will be performed, how will it be “looked” at, what will be calculated. Do it smart and you can have a predictable/controllable lossy compression with seriously small file sizes, do it without thinking and your data will lead to misinterpretation and apocalypse.

Von wo wurde der Hamburger Fernsehturm fotografiert

Vom großartigen @Fernsehturm_HH-Account inspiriert hab ich mal ein 4 Jahre altes Projektchen aufgewärmt: Verortete Fotos auf Flickr, die mit “Hamburg” und “Fernsehturm” getaggt sind. Alternativ als Linien vom (angeblichen) Aufnahmeort zum Turm.

Image

Ungefiltert und voller Mist, aber man kann z. B. toll sehen, wie einladend die Brückem vorm Kuhmühlenteich für Fotos sind.

Kostenlose Internetpunkteabsahnidee: Dasselbe mit dem Eiffelturm, dem Londoner Dildo u. ä. machen, Konkave Hülle außen drum, schick aufbereiten, als “Nur wo man $Wahrzeichen sieht, ist $Stadt” vermarkten, €€€€.

Highlight current timeslice in a QGIS Atlas layout

Did this for an ex-colleague some months ago and forgot to share the how-to publically. We needed a visual representation of the current time in a layout that showed both a raster map (different layer per timeslice) and a timeseries plot of an aspect of the data (this was created outside QGIS).

Have lots of raster layers you want to iterate through. I have:

./ECMWF_ERA_40_subset/2019-01-01.tif
./ECMWF_ERA_40_subset/2019-01-02.tif
./ECMWF_ERA_40_subset/2019-01-03.tif
./ECMWF_ERA_40_subset/2019-01-04.tif
./ECMWF_ERA_40_subset/2019-01-05.tif
./ECMWF_ERA_40_subset/2019-01-06.tif
...

Create a new layer for your map extent. Draw your extent as geometry. Duplicate that geometry as many times as you have days. Alternatively you could of course have different geometries per day. Whatever you do, you need a layer with one feature per timeslice for the Atlas to iterate though. I have 30 days to visualise so I duplicated my extent 30 times.

Open the Field Calculator. Add a new field called date as string type (not as date type until some bug is fixed (sorry, did not make a note here, maybe sorting is/was broken?)) with an expression that represents time and orders chronologically if sorted by QGIS. For example: '2019-01-' || lpad(@row_number,2,0) (assuming your records are in the correct order if you have different geometries…)

Have your raster layers named the same way as the date attribute values.

Make a new layout.

For your Layout map check “Lock layers” and use date as expression for the “Lock layers” override. This will now select the appropriate raster layer, based on the attribute value <-> layer name, to display for each Atlas page.

Cool, if you preview the Atlas now you got a nice animation through your raster layers. Let’s do part 2:

In your layout add your timeseries graph. Give it a unique ID, e. g. “plot box”. Set its width and height via new variables (until you can get those via an expression this is needed for calculations below).

Create a box to visualise the timeslice. Set its width to map_get(item_variables('plot box'), 'plot_width') / @atlas_totalfeatures. For the height and y use/adjust this expression: map_get(item_variables('plot box'), 'plot_height'). For x comes the magic:

with_variable(
'days_total',
day(to_date(maximum("date"))-to_date(minimum("date")))+1,
-- number of days in timespan
-- +1 because we need the number of days in total
-- not the inbetween, day() to just get the number of days
with_variable(
'mm_per_day',
map_get(item_variables('plot box'), 'plot_width') / @days_total,
with_variable(
'days',
day(to_date(attribute(@atlas_feature, 'date'))-to_date(minimum("date"))),
-- number of days the current feature is from the first day
-- to_date because BUG attribute() returns datetime for date field
@mm_per_day * @days + map_get(item_variables('plot box'), 'plot_x')
)
)
)

This will move the box along the x axis accordingly.

Have fun!

#30DayMapChallenge: Day 3 – Polygons OR Lego® style brick raster in QGIS using Geometry Generator expressions

I showed this at the annual meeting of the swiss user group in June 2019 and promptly forgot to post it for everyone to see. Let’s blame the scenery?

Have a raster.

Processing -> Create Grid, covering it with points in a spacing of your choice. Use the same CRS as your raster (unless you want to figure out expression-based geometry transformations on your own, like in my elevation lines code)

Change the Symbol layer type to Geometry Generator and enter

with_variable(
  'radius',
  3333,
  buffer($geometry, @radius, 16)
)

where radius should be a value about one third of your spacing.

You should see circles!

For the fill color use this expression and adjust the name of your raster layer:

with_variable(
    'raster_layer',
    'DHM200.xyz',
    ramp_color('Blues',  -- change to an other named ramp here if you like
        scale_linear(
            raster_value(
                @raster_layer, 1, 
                centroid($geometry)  -- back to our point
            ),
            raster_statistic(
                @raster_layer, 1, 
                'min'
            ),
            raster_statistic(
                @raster_layer, 1, 
                'max'
            ),
            0, 1  -- new scale as color ramps go 0 to 1
        )
    )
)

This will get the raster value below our grid point and fit it onto a color ramp between the min and max of all the raster values.

For the stroke use the same expression but wrap darker(..., 150) around it so you get a darker color.

Using Draw Effects add a small Drop Shadow to your circles. I used an offset of 0.5 mm and a blur radius of 1 mm.

Now add another Geometry Generator symbol layer below your existing one and use the following expression:

bounds(buffer($geometry, 5000))

with_variable(
  'radius',
  5000,
  bounds(
    buffer($geometry, @radius)
  )
)

with the radius being half your grid spacing.

Use the same expressions for the colors as above but set the darker value to 200.

For some more fanciness maybe add a “QGIS” text on top of the nupsies?

Exercise: Make it so that the result perfectly covers the raster, instead of being one grid cell off like mine.

Thanks Topi.

#30DayMapChallenge: Day 2 – Lines

Thanks Topi.

Addresses in Hamburg: http://suche.transparenz.hamburg.de/dataset/alkis-adressen-hamburg15

Convert the housenumber from string to integer: to_int("hausnr")

Processing -> Points to Path using above housenumber as order field and strname as group field.

Color those ziggyzaggety lines as you like.

Maybe explode the lines and style each segment by its housenumber?

And put them on an interactive slippy map.

Dynamic elevation profile lines as QGIS geometry generator

Load a raster layer in QGIS.

Add a new Scratch Layer of Polygon type (in any CRS). Set its Symbology to Inverted Polygons. Use a Geometry Generator as symbol layer type. Set it to LineString/MultiLineString. Enter the expression below and adjust the layer ID (best enter the expression editor and find it in the “layers” tree). Then adjust the scaling factor at the very bottom.

-- UPPER CASE comments below are where you can change things

with_variable(
	'raster_layer',
	'long_and_complicated_layer_id',  -- RASTER LAYER to sample from
	-- this collects all the linestrings generated below into one multilinestring
	collect_geometries(
		-- a loop for each y value of the grid
		array_foreach(
			-- array_foreach loops over all elements of the series generated below
			-- which is a range of numbers from the bottom to the top of y values
			-- of the map canvas extent coordinates.
			-- the result will be an array of linestrings
			generate_series(
				y(@map_extent_center)-(@map_extent_height/2), -- bottom y
				y(@map_extent_center)+(@map_extent_height/2),  -- top y
				@map_extent_height/50  -- stepsize -> HOW MANY LINES
			),
			
			-- we want to enter another loop so we assign the name 'y' to
			-- the current element of the array_foreach loop
			with_variable(
				'y',
				@element,
				
				-- now we are ready to generate the line for this y value
				make_line(
					-- another loop, this time for the x values. same logic as before
					-- the result will be an array of points
					array_foreach(
						generate_series(
							x(@map_extent_center)-(@map_extent_width/2), -- left x
							x(@map_extent_center)+(@map_extent_width/2),  -- right x
							@map_extent_width/50  -- stepsize -> HOW MANY POINTS PER LINE
						),
						-- and here we create each point of the line
						make_point(
							@element,  -- the current value from the loop over the x value range
							@y  -- the y value from the outer loop
							+   -- will get an additional offset to generate the effect
							-- we look for values at _this point_ in the raster, and since
							-- the raster might not have any value here, we must use coalesce
							-- to use a replacement value in those cases
							coalesce(  -- coalesce to catch raster null values
								raster_value(
									@raster_layer,
									1,  -- band 1, *snore*
									-- to look up the raster value we need to look in the right position
									-- so we make a sampling point in the same CRS as the raster layer
									transform( 
										make_point(@element, @y),
										@map_crs,
										layer_property(@raster_layer,'crs')
									)
								),
								0  -- coalesce 0 if raster_value gave null
							-- here is where we set the scaling factor for the raster -> y values
							-- if things are weird, set it to 0 and try small multiplications or divisions
							-- to see what happens.
							-- for metric systems you will want to multiply
							-- for geographic coordinates you will want to divide
							)*10  -- user-defined factor for VERTICAL EXAGGERATION
						)
					)
				)
			)
		)
	)
)  -- wee

If you don’t have your raster data on a SSD this can be a bit slow.

Yes, this works if you change your CRS!

Anaglyph 3D contour lines

A slight misunderstanding about a weird pattern I posted to Twitter earlier made me wonder how easy it might be to present elevation data with real depth ™ as anaglyph 3D in QGIS. Anaglyph 3D, you know, those red and cyan glasses that rarely ever worked.

You might remember my post on fake chromatic aberration and dynamic label shadows in QGIS some months ago. Maybe with some small adjustments of the technique…?

Wikipedia has an detailed article on the topic if you are interested. It boils down to: One eye gets to see not the red but the cyan stuff. The other sees the red but not the cyan. By copying, coloring and shifting elements left-right increasingly the human vision gets tricked into seeing fake depth.

Got your anaglyph glasses? Oogled some images you found online to get in the mood? Excellent, let’s go!

Get some DEM data. I used a 10 meter DEM of Hamburg and a 200 meter DEM of Germany for testing. Make sure both your data and your project/canvas CRS match or you have a hard time with the math.

Extract contours at a reasonable interval. What’s reasonable depends on your data. I used an interval of 5 meters for the 10 meter DEM as Hamburg is rather flat (which you will NOT see here).

You may want to simplify or smooth this.

Now it’s time to zoom in, because the effect only works if the line density is appropriately light.

Now we need to duplicate those, color them and move them left and right.

Change the symbol layer of your contours to a Geometry Generator.

Let’s use that first layer for the left, red part. So change the color to red. Use a red that vanishes when you look through your left (red) eye but is clearly visible through the right (cyan) eye. The exact color depends on your glasses.

Set the Geometry Generator to LineString. I will now explain an intermediate step so if you just want the result, scroll a bit.

translate(
  $geometry,
  -  -- there is a minus here!
  (
    x_max(@map_extent) - x(centroid($geometry))
  )
  /100  -- magic scale value…
  ,
  0  -- no translation in y
)

This moves each contour line to the left for a value that increases with the geometry’s distance to the right side. Since we don’t want to move the geometries too far, a magic scale factor needs to be added and adjusted according to your coordinate values.

(Yes, that is a bug right here (centroid is a bad metric and some of the contours are huge geometries) but hey it works well enough. Segmentizing could fix this. Or just extract the vertices, that looks cool too TODO imagelink)

For the right, cyan side we need to add another Geometry Generator symbol layer, color it in cyan (so that you can only see it through your left eye) and do the geometry expression the other way around:

translate(
  $geometry,
  (
    x(centroid($geometry))-x_min(@map_extent)
  )
  /100  -- magic scale value…
  ,
  0  -- no translation in y
)

Cool! But this is lacking a certain depth, don’t you think? We need to scale the amount of horizontal shift by the elevation value of the contour lines. And then we also need to adjust the magic scale value again because reasons.

For the red symbol layer:

translate(
  $geometry,
  -- move to the LEFT
  -- scaled by the distance to the RIGHT side of the map
  -- scaled by the elevation

  -- minus so it goes to the left
  -
  "ELEV" -- the attribute field with the elevation values
  *
  (x_max(@map_extent)-x(centroid($geometry)))
  
  -- MAGIC scale value according to CRS and whatever
  /10000
  ,
  -- no change for y
  0
)

For the cyan symbol layer:

translate(
  $geometry,
  -- move to the RIGHT
  -- scaled by the distance to the LEFT side of the map
  -- scaled by the elevation
 
  "ELEV" -- the attribute field with the elevation values
  *
  (x(centroid($geometry))-x_min(@map_extent))
  
  -- MAGIC scale value according to CRS and whatever
  /10000
  ,
  -- no change for y
  0
)

That’s it!

Now, who is going to do Autostereograms? Also check out http://elasticterrain.xyz.