Fix the ITS_LIVE tutorial

docs/src/tutorials/its_live.jl

@@ -2,12 +2,21 @@
 # ITS_LIVE data
 =#
 
-using Rasters, ZarrDatasets
-using Kerchunk
-using Statistics
-
-rs = RasterStack("reference://$(joinpath(dirname(dirname(pathof(Kerchunk))), "test", "data", "its_live_catalog.json"))")
-#
+using Rasters       # Raster data analysis in Julia
+using ZarrDatasets  # Zarr support for Rasters
+using Kerchunk      # Kerchunk support for Zarr
+using Statistics    # Basic statistics
+# We can load the catalog from the catalog file that we ship in the repo:
+catalog_path = joinpath(dirname(dirname(pathof(Kerchunk))), "test", "data", "its_live_catalog.json")
+rs = RasterStack("reference://$(catalog_path)"; source = Rasters.Zarrsource())
+# We've now loaded the dataset lazily in a `RasterStack`, which is essentially a stack of multiple variables.
+# Now, we can apply arbitrary Rasters.jl functions to the stack, or plot it, and treat it as a general Julia array!
+# 
+# Let's plot first:
+using CairoMakie
+heatmap(rs.v)
+# We can also aggregate the data to a lower resolution, which downloads the entire dataset.
+# Here, we aggregate by a factor of 10 in both dimensions, so a 10x10 window is aggregated to a single pixel.
 vs2 = Rasters.aggregate(rs, mean, 10) # now everything is loaded in disk
-#
+# and plot this aggregated data:
 arrows(dims(vs2, X) |> collect, dims(vs2, Y) |> collect, vs2.vx .* 20, vs2.vy .* 20; arrowsize = 5)