Building RASP
If you want to generate Blipmap Forecasts by yourself, using RASP from scratch is a good starting point. It allows you to build a docker image to generate forecasts. But there are a few things you need to do to build this docker image. This post explains how you can build the simplest docker image from it on AWS EC2.
Creating an EC2 instance
First, let’s create an EC2 instance. I created m7i.xlarge image using the latest Ubuntu 24.04 LTS AMI. I recommend you attach an EBS larger than 200GB to it. Once you’ve created an instance, login to it and run
sudo apt-get update
sudo apt-get upgrade
to update everything.
Installing docker
Next, let’s install docker.
sudo apt-get install ca-certificates curl
sudo install -m 0755 -d /etc/apt/keyrings
sudo curl -fsSL https://download.docker.com/linux/ubuntu/gpg -o /etc/apt/keyrings/docker.asc
sudo chmod a+r /etc/apt/keyrings/docker.asc
echo \
"deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.asc] https://download.docker.com/linux/ubuntu \
$(. /etc/os-release && echo "${UBUNTU_CODENAME:-$VERSION_CODENAME}") stable" | \
sudo tee /etc/apt/sources.list.d/docker.list > /dev/null
sudo apt-get update
sudo apt-get install docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin
sudo usermod -aG docker ubuntu
Once you’ve done this, reboot your instance.
Building docker images
You’re now ready to build docker images. Clone the repo and build base, wrf_build and wrf_prod images.
git clone https://github.com/sfalmo/rasp-from-scratch.git
cd rasp-from-scratch
cp .env.template .env
You can optionally update .env to use newer versions of WRF and WPS.
--- a/.env 2025-09-28 08:25:48.211582952 +0000
+++ b/.env 2025-10-06 12:56:02.602479862 +0000
@@ -7,8 +7,8 @@
RASP_IMAGE=your.docker.repo/aufwinde/rasp/rasp
# Versions of WRF and WPS that shall be used. Those must be valid tags in the corresponding GitHub repo
-WRF_VERSION=v4.5
-WPS_VERSION=v4.5
+WRF_VERSION=v4.7.1
+WPS_VERSION=v4.6.0
# Choose the CPU architecture for the machine you are currently using for the build (native should be fine)
WRF_MARCH_BUILD=native
Once you’ve updated (or not updated) .env, you’re ready to build these three images.
docker compose build base
docker compose build wrf_build
docker compose build wrf_prod
Before building rasp image, we need to prepare geographical data. The repo is configured to use some high resolution data, but we’ll use the simplest ones we can download from WPS V4 Geographical Static Data Downloads Page by patching namelist.wps.
diff --git a/rasp/TIR/namelist.wps b/rasp/TIR/namelist.wps
index 4ea4532..ad755ae 100755
--- a/rasp/TIR/namelist.wps
+++ b/rasp/TIR/namelist.wps
@@ -14,7 +14,7 @@
j_parent_start = 1, 25,
e_we = 90, 206,
e_sn = 90, 206,
- geog_data_res = 'gmted2010_30s+corine_usgs_500m+bnu_soil_30s+modis_fpar+modis_lai+2m', 'SRTM+corine_usgs_250m+bnu_soil_30s+modis_fpar+modis_lai+30s',
+ geog_data_res = 'default', 'default',
dx = 10000,
dy = 10000,
map_proj = 'lambert',
Also, we need to patch namelist.input because these data don’t contain lake physics data.
diff --git a/rasp/TIR/namelist.input b/rasp/TIR/namelist.input
index 8a98b1e..6c18c47 100755
--- a/rasp/TIR/namelist.input
+++ b/rasp/TIR/namelist.input
@@ -84,7 +84,7 @@
sf_sfclay_physics = 1, 1, 1,
sf_surface_physics = 2, 2, 2,
sf_urban_physics = 0, 0, 0,
- sf_lake_physics = 1, 1, 1,
+ sf_lake_physics = 0, 0, 0,
bl_pbl_physics = 1, 1, 1,
bldt = 0, 0, 0,
topo_wind = 1, 1, 1,
Let’s prepare our geographic data file. We’ll download some archives and put them together into geog.tar.gz.
curl -o geog_high_res_mandatory.tar.gz https://www2.mmm.ucar.edu/wrf/src/wps_files/geog_high_res_mandatory.tar.gz
curl -o landuse_30s_with_lakes.tar.bz2 https://www2.mmm.ucar.edu/wrf/src/wps_files/landuse_30s_with_lakes.tar.bz2
mkdir geog
(cd geog && tar zxf ../geog_high_res_mandatory.tar.gz --strip-components=1)
(cd geog && tar jxf ../landuse_30s_with_lakes.tar.bz2)
tar zcf rasp/geog.tar.gz geog
Once you’ve prepared geog.tar.gz and patched these files, you’re ready to build rasp image.
docker compose build rasp
You can run this image now to build your forecasts.
docker compose run rasp
The results will be stored in results/OUT directory in your instance. Look at results/LOG/GM.stderr to find what was wrong when it failed.
Sometimes, building a docker image doesn’t fail even when geogrid.exe fails. In this case, you’ll find an error in results/LOG/metgrid.out. Run bash using the image to investigate it further.
docker compose run rasp bash
You’ll find why geogrid.exe failed by checking TIR/geogrid.out in the container.