Optical / multispectral — Sentinel-2 SR Harmonized¶
Sentinel-2 Surface Reflectance Harmonized — June 2024 median of the red band (B4) over Giza at 30 m. Median reduction is the canonical cloud-screened composite.
Setup¶
Imports and a per-notebook output directory. pyramids provides Dataset (GeoTIFF reading + plotting); earthlens provides the unified EarthLens entry point and the GEE Catalog.
import os
from pathlib import Path
import matplotlib.pyplot as plt
import numpy as np
from pyramids.dataset import Dataset as PyramidsDataset
from earthlens import EarthLens
from earthlens.gee import Catalog
OUT_DIR = Path('out') / 'optical-multispectral'
OUT_DIR.mkdir(parents=True, exist_ok=True)
print(f'output directory: {OUT_DIR.resolve()}')
Credentials¶
The notebook reads the GEE service-account credentials from the GEE_SERVICE_ACCOUNT / GEE_SERVICE_KEY environment variables. Both must be set before running the download cells below.
SERVICE_ACCOUNT = os.environ['GEE_SERVICE_ACCOUNT']
SERVICE_KEY = os.environ['GEE_SERVICE_KEY']
Inspect the catalog entry¶
Before downloading anything, look at what the bundled catalog knows about the asset — bands, cadence, license, provider.
cat = Catalog()
ds = cat.get_dataset('COPERNICUS/S2_SR_HARMONIZED')
print(f'title: {ds.title}')
print(f'ee_type: {ds.ee_type}')
print(f'spatial_resolution: {ds.spatial_resolution} m')
print(f'extent.start_date: {ds.extent.start_date}')
print(f'extent.end_date: {ds.extent.end_date}')
print(f'default_reducer: {ds.default_reducer}')
print(f'license: {ds.license}')
print(f'provider: {ds.provider}')
print(f'#bands: {len(ds.bands)}')
print(f'band ids (first 5): {list(ds.bands)[:5]}')
Download¶
Tiny AOI ([29.95, 30.05] lat, [31.15, 31.25] lon) at 30.0 m, monthly cadence — keeps the synchronous download under EE's 32768-px per-axis cap. The request and authentication are kept on separate lines so each step is easy to read and re-run.
gee = EarthLens(
data_source="gee",
start='2024-06-01',
end='2024-06-30',
dataset='COPERNICUS/S2_SR_HARMONIZED',
variables=['B4'],
aoi=[31.15, 29.95, 31.25, 30.05],
cadence='monthly',
path=str(OUT_DIR),
scale=30.0,
reducer='median',
)
gee.authenticate(service_account=SERVICE_ACCOUNT, service_key=SERVICE_KEY)
Run the synchronous download. The try/except keeps the graceful-skip toggle from the original notebook (currently disabled via if False:), so a live EE failure re-raises by default; download_ok records whether the GeoTIFF was written.
try:
paths = gee.download(progress_bar=False)
print(f'wrote {len(paths)} GeoTIFF(s):')
for p in paths:
print(f' {p} ({p.stat().st_size / 1024:.1f} KB)')
download_ok = True
except Exception as exc:
if False:
print(
f'live EE call failed (tolerated for this category): {type(exc).__name__}: {exc}'
)
paths = []
download_ok = False
else:
raise
Quick preview¶
Load the first written GeoTIFF through pyramids and render the single band. (pyramids.dataset.Dataset is the project's GeoTIFF/NetCDF wrapper.) The dataset's nodata is masked so the colormap isn't pinned to it.
if not download_ok or not paths:
print('Skipping preview — no GeoTIFF was written.')
else:
pds = PyramidsDataset.read_file(str(paths[0]))
arr = pds.read_array()
if arr.ndim == 3:
arr = arr[0]
# Mask the dataset's nodata so the colormap doesn't get pinned to it.
nodata = pds.no_data_value
if nodata is not None:
try:
arr = np.ma.masked_equal(
arr, float(nodata[0] if isinstance(nodata, (list, tuple)) else nodata)
)
except (TypeError, ValueError):
pass
fig, ax = plt.subplots(figsize=(6, 5))
im = ax.imshow(arr, cmap='viridis')
ax.set_title(f'{"COPERNICUS/S2_SR_HARMONIZED"} / {"B4"}')
ax.set_xlabel('x (px)')
ax.set_ylabel('y (px)')
fig.colorbar(im, ax=ax, shrink=0.8)
plt.tight_layout()
plt.show()
print(f'value range: [{float(np.nanmin(arr)):.4g}, {float(np.nanmax(arr)):.4g}]')
Tracking submitted jobs (asynchronous export)¶
The download above uses export_via="url" — a synchronous getDownloadURL round-trip. Nothing was queued, so there's no Earth Engine job to track.
To track an export instead, switch to an asynchronous sink (drive / gcs / asset) and pass wait_for_export=False so .download() returns a TaskInfo at submission time rather than blocking until completion. The cells below submit the same (asset_id, band, AOI, scale) request as an export_via="asset" task into the service account's own asset folder, then walk the four jobs-API calls (list_recent_tasks → wait_for_task_id → ee.data.getAsset → ee.data.deleteAsset) to make the job finish and tidy up. See track-batch-exports.ipynb for a deeper worked example.
Prepare the demo asset folder¶
GEE._export_via_batch writes the image at <asset_id>/<prefix>, so asset_id here is the parent folder, not the final image path. Clear any leftover children from a previous run, then create the empty folder EE requires before a child write.
import ee
from earthlens.gee import cancel_task, list_recent_tasks, wait_for_task_id
# The asset goes into a `Folder` asset that we own. `GEE._export_via_batch`
# writes the actual image at `<asset_id>/<prefix>`, so `asset_id` here is
# the parent FOLDER (not the final image path). Both must be cleaned up.
_proj = ee.data._get_projects_path().removeprefix('projects/')
DEMO_FOLDER = f'projects/{_proj}/assets/earthlens-demo-optical-multispectral'
print(f'demo folder: {DEMO_FOLDER}')
# Best-effort cleanup of leftover children from a previous run (so the
# folder is empty before we try to delete it below).
try:
for child in ee.data.listAssets({'parent': DEMO_FOLDER}).get('assets', []):
ee.data.deleteAsset(child['name'])
print(f'cleared leftover child: {child["name"]}')
ee.data.deleteAsset(DEMO_FOLDER)
print(f'cleared leftover folder: {DEMO_FOLDER}')
except Exception:
pass
# Create the parent folder — EE requires it to exist before a child write.
ee.data.createAsset({'type': 'Folder'}, DEMO_FOLDER)
print(f'created folder: {DEMO_FOLDER}')
Submit¶
Same (asset_id, band, AOI, scale) request as the sync download above, just routed through export_via="asset" + wait_for_export=False. The construct → authenticate → download steps are kept on separate lines; download() returns a TaskInfo per submitted bucket at the moment the task is queued — no blocking.
submitted_ok = False
task_info = None
try:
async_gee = EarthLens(
data_source="gee",
start='2024-06-01',
end='2024-06-30',
dataset='COPERNICUS/S2_SR_HARMONIZED',
variables=['B4'],
aoi=[31.15, 29.95, 31.25, 30.05],
cadence='monthly',
path=str(OUT_DIR),
scale=30.0,
reducer='median',
export_via='asset',
asset_id=DEMO_FOLDER,
wait_for_export=False,
)
async_gee.authenticate(service_account=SERVICE_ACCOUNT, service_key=SERVICE_KEY)
submitted = async_gee.download(progress_bar=False)
task_info = submitted[0]
submitted_ok = True
print(f'submitted: id={task_info.id} state={task_info.state}')
print(f' description={task_info.description}')
except Exception as exc:
if False:
print(f'async submission failed (tolerated): {type(exc).__name__}: {exc}')
else:
raise
List + wait¶
list_recent_tasks(description_prefix=...) returns every matching task across the current project; wait_for_task_id blocks until the one we care about reaches a terminal state. A real workflow would just poll later from a separate process — the wait here exists so the notebook shows the full success path end-to-end.
if submitted_ok and task_info is not None:
recent = list_recent_tasks(
description_prefix=task_info.description,
max_age_min=10,
)
print(f'list_recent_tasks matched {len(recent)} task(s):')
for t in recent:
print(f' {t.id} {t.state:<12} {t.description}')
try:
final = wait_for_task_id(
task_info.id,
poll_seconds=10,
progress_bar=False,
)
print(f'\nfinal state: {final.state}')
except RuntimeError as exc:
# Raised on FAILED / CANCELLED — cancel-if-still-running
# so we don't leak an in-flight task on notebook restart.
print(f'wait_for_task_id raised: {exc}')
try:
cancel_task(task_info.id)
except Exception:
pass
else:
print('Skipping list/wait — async submission did not succeed.')
Verify + clean up¶
Confirm the produced asset exists on Earth Engine, then delete it (and the surrounding demo folder) so we don't leak storage between notebook runs. The backend wrote the image at <DEMO_FOLDER>/<task description>.
if submitted_ok and task_info is not None:
produced = f'{DEMO_FOLDER}/{task_info.description}'
try:
meta = ee.data.getAsset(produced)
print(f'asset exists: type={meta.get("type")} ' f'name={meta.get("name")}')
ee.data.deleteAsset(produced)
print('asset deleted')
except Exception as exc:
print(f'verify/delete skipped: {exc}')
# Always try to tear down the parent folder.
try:
ee.data.deleteAsset(DEMO_FOLDER)
print(f'folder deleted: {DEMO_FOLDER}')
except Exception as exc:
print(f'folder delete skipped: {exc}')
What's on disk¶
The GeoTIFF (or empty list, if the EE call was tolerated) is left under the per-notebook out/ directory for you to inspect. That directory is .gitignored — re-running the notebook overwrites it.
for p in sorted(OUT_DIR.iterdir()) if OUT_DIR.exists() else []:
print(f'{p} ({p.stat().st_size / 1024:.1f} KB)')