ScientistCloud GDAL-to-IDX Dashboard

This standalone notebook downloads terrain-parameter GeoTIFF files from the public ScientistCloud CONUS datasets, reads them with GDAL, converts selected fields to an OpenVisus IDX dataset, validates the result, and launches the OpenVisus dashboard.

Default example:

• Resolution: 30m
• State: TN
• Fields: elevation, hillshade, aspect, slope, plan_curvature

Setup

Python 3.10 is recommended. Install GDAL with Conda, then install the Python requirements with pip:

conda create -n scientistcloud-idx python=3.10 gdal=3.8.4 pip
conda activate scientistcloud-idx
pip install -r requirements.txt

If you must use pip-only installation, install a GDAL wheel compatible with your system before running the notebook. GDAL installed only through pip can be fragile because it depends on native GDAL libraries.

If you installed the requirements before param==2.0.2 was pinned, refresh the environment with:

pip install --upgrade --force-reinstall -r requirements.txt

To use the environment in Jupyter:

python -m pip install ipykernel jupyterlab
python -m ipykernel install --user --name scientistcloud-idx --display-name "ScientistCloud IDX"
jupyter lab

Open Transform_ScientistCloud_to_IDX.ipynb and select the ScientistCloud IDX kernel.

Running The Notebook

The notebook stores files locally under:

• TIFF downloads: data/tif/<resolution>/<state>/
• IDX output: data/idx/
• Dashboard log: logs/dashboard.log

Run the cells in order. The dashboard cell is intentionally short:

%%capture dashboard_output
launch_dashboard()

After the dashboard starts, open:

http://localhost:8989/dashboard

Use localhost, not 0.0.0.0, in your browser.

Configuration

Edit the configuration cell near the top of the notebook:

RESOLUTION = "30m"
STATE = "TN"
FIELDS = ["elevation", "hillshade", "aspect", "slope", "plan_curvature"]
SOURCE_MODE = "local_or_download"

Supported resolutions are:

• 30m
• 10m

The 10m files can be many gigabytes per field. Make sure you have enough disk space before switching to 10m or adding more fields.

SOURCE_MODE options:

• local_or_download: reuse local TIFFs when present, otherwise download them.
• download: download configured TIFFs even if local files already exist.
• local: require TIFFs to already exist locally.

Recommended Dashboard Ranges

If the dashboard only shows the silhouette of the state, the color range is probably including the NoData value -999999. Change Range from dynamic to a manual/user range and use these starting values:

Field	Min	Max
elevation	0	2000
hillshade	0	255
aspect	0	360
slope	0	70
plan_curvature	-0.05	0.05

For terrain interpretation, hillshade and slope usually show spatial structure most clearly at first glance. For elevation, keep Mapper as linear and use a palette such as Viridis256, Turbo256, or Inferno256.

Troubleshooting

If the dashboard cell exits immediately, run the diagnostic cell:

dashboard_output.show()

Also check:

logs/dashboard.log

If you see NumPy/GDAL compatibility errors, confirm that the environment uses numpy<2:

python -c "import numpy; print(numpy.__version__)"

If openvisuspy or OpenVisus cannot be imported, rerun:

pip install -r requirements.txt