Add V2 example derived from PO.DAAC tutorial

examples/V2/ccmp/.gitignore

@@ -0,0 +1 @@
+ccmp_icechunk_*

examples/V2/ccmp/README.md

@@ -0,0 +1,45 @@
+# CCMP Recipes: VirtualiZarr + Icechunk with NASA Earthdata
+
+Notebook demonstrating how to virtualize the ~750 GB CCMP wind dataset from PO.DAAC, write to Icechunk, parallelize with Dask, and append new data. Based on the Po.DAAC tutorial by Dean Henze at https://podaac.github.io/tutorials/notebooks/Advanced_cloud/virtualizarr_recipes.html.
+
+## Prerequisites
+
+- NASA Earthdata account: https://urs.earthdata.nasa.gov/
+- Credentials via `EARTHDATA_USERNAME`/`EARTHDATA_PASSWORD` env vars or `~/.netrc`
+- Must run from **AWS us-west-2** for direct S3 access
+- Recommended: `m6i.4xlarge` EC2 instance (16 CPUs, 64 GiB) for parallel sections
+
+## Setup
+
+Install [uv](https://docs.astral.sh/uv/) if you haven't already.
+
+Sync the project (creates a `.venv` and installs all dependencies):
+
+```bash
+uv sync
+```
+
+
+## Register the Jupyter kernelspec
+
+Register the project's virtualenv as a Jupyter kernel so you can select it in JupyterLab:
+
+```bash
+uv run python -m ipykernel install --user --name ccmp-recipes --display-name "CCMP Recipes"
+```
+
+Then launch JupyterLab (from any environment that has it installed) and select the **CCMP Recipes** kernel when opening the notebook.
+
+## Run
+
+Open the notebook in JupyterLab and select the `ccmp-recipes` kernel:
+
+```bash
+uv run jupyter lab
+```
+
+Or run it non-interactively:
+
+```bash
+uv run jupyter execute virtualizarr_recipes.ipynb
+```