• read_lidar
• calc_heights_from_lidar

Del PNOA:

• Nubes de puntos (resolución >= 0.5 puntos/m2). 1ª cobertura: 2009-2015. 2ª cobertura (aún no disponible para todo el territorio): 2015-2020.

• Modelo Digital del Terreno: altura del suelo, generado por interpolación de puntos clase 2 (terreno), con resolución de 5 m (1ª cobertura) o 2 m (2ª cobertura).

• Modelo Digital de Superficies: alturas sobre el terreno, generado a partir de puntos de clase 2 (terreno), 3, 4, 5 (vegetación) y 6 (edificaciones), con resolución de 5 m (1ª cobertura). También disponibles MDS de vegetación y edificaciones (por separado), con resolución de 2.5 m (1ª cobertura).

• Mapa Lidar: Modelo Digital de Superficies sombreado, distinguiendo vegetación de edificaciones, con resolución de 2.5 m

OSM Buildings

(Catastro tiene información sobre número de plantas, pero no alturas)

Time zones together with Daylight Savings Times are complicated...

The time difference between local time and UTC/solar time at late March or late October (i.e. after DST) depends on the particular year

We're going to model a 'standard' year, so let's assume that all days from November to March, both included, have winter time (i.e. 1 hour ahead of UTC/solar), and all days between April and October (both included) have summer time (i.e. 2 hours ahead of UTC/solar)

P.S. Here is a base R solution to find time difference at any given date

Could be more useful/accurate in our case. See end of https://www.tylermw.com/throwing-shade/

Remove pixels occupied by buildings. Probably water too
That is, leave pixels that are outdoors (streets, open spaces), either vegetated or not, to calculate shade statistics across districts, etc
We can do this using lidar points classes (see #2 (comment))
Vegetation <1m should probably be discarded

Coming from #8

We need to discount the number of rays that are intercepted by the tree canopy just above, but keep all the other rays (coming from sides).

Look into rayshader::ray_shade function and try to get the number of rays intercepted vertically by the canopy above.

See also explanation of algorithm here https://www.tylermw.com/throwing-shade/

• single small tree

• single big tree (same as several small trees together)

• square building with open area inside

This can be useful to prototype the shiny app: https://ashbaldry.github.io/designer/

Writing functions: https://rstudio.cloud/learn/primers/6

Building R packages: https://r-pkgs.org/

Advanced R: https://adv-r.hadley.nz/

We need to find the best route between two points (ideally, two addresses or geoplaces as in Google Maps or Openstreetmap) with the conditions that

• route must follow streets (ideally from OSM), i.e. not being a direct line between the two points. Must be possible to make it walking

• route must minimize cost, in terms of selecting the most shaded routes as coming from the shade raster for that date and time

For that, we will use lidR::normalize_height, then lidR::rasterize_canopy, then reclassify those heights into ground level or not

I want to get a raster with the cover type (i.e. building, vegetation, water) that comes with the LiDAR points data

@jesus-davila do you know an easy way to put that information (ie. LiDAR point classes) on a raster? Thanks

library(terra)
#> terra 1.5.21
library(raster)
#> Loading required package: sp

r = rast(nrows = 3, ncols = 3, xmin = 0, xmax = 3, ymin = 0, ymax = 3, 
         resolution = 1, vals = 1:9)
plot(r)

raster <- raster(r)

m <- matrix(raster::extract(raster, raster::extent(raster)), 
              nrow = ncol(raster), ncol = nrow(raster))

m
#>      [,1] [,2] [,3]
#> [1,]    1    4    7
#> [2,]    2    5    8
#> [3,]    3    6    9

See https://github.com/tylermorganwall/rayshader/blob/7d19e315c8f171f88bc11eb30f7c464ebc672cdd/R/raster_to_matrix.R#L33

Need to transpose resulting matrix in terra_to_matrix

Generate SpatRaster with many layers