HarmonyPy Tutorial

Harmony Py Tutorial¶

This notebook shows three basic examples of Harmony jobs, each using a Harmony test Collection. The first example requests a spatial subset of Alaska, the second a temporal subset (a single-month timespan), and the third shows a combination of both spatial and temporal subsetting.

First, we import a few things that will help us create a request and display images. We then import the Harmony Py classes we need to make a request.

In [1]:

import sys; sys.path.append('..')
import datetime as dt
from IPython.display import display, JSON
import rasterio
import rasterio.plot

from harmony import BBox, Client, Collection, Request

Now we create a Harmony Client object, letting it pick up our credentials from a .netrc file.

In [2]:

harmony_client = Client()

Next, let's create a Collection object with the CMR collection id for the CMR collection we'd like to look at.

We then create a Request which specifies that collection, a spatial BBox describing the bounding box for the area we're interested in (we'll look at the BBox in other tutorials). In this case we're interested in looking at Alaska (and who wouldn't be?). We also include a date/time range to narrow down the data.

Because this data includes a lot of different variables, we limit it by passing in a list of variables we're interested in; in this test collection we'll look at the blue variable. We include a max_results parameter to limit the results to the first 10 images just to get a sample of what things look like.

Next, we include a coordinate reference system (CRS) indicating we'd like to reproject the data into the Arctic Polar Stereographic projection. We also specify that we'd like the output to be in the GeoTIFF format with a resolution of 512x512 pixels.

Finally we check if the request we've created is valid.

In [3]:

collection = Collection(id='C1234088182-EEDTEST')

request = Request(
    collection=collection,
    spatial=BBox(-165, 52, -140, 77),
    temporal={
        'start': dt.datetime(2010, 1, 1),
        'stop': dt.datetime(2020, 12, 30)
    },
    variables=['blue_var'],
    max_results=10,
    crs='EPSG:3995',
    format='image/tiff',
    height=512,
    width=512
)

request.is_valid()

Out[3]:

True

Now that we have a request, we can submit it to Harmony using the Harmony Client object we created earlier. We'll get back an id for our request which we can use to find the job's status and get the results.

In [4]:

job_id = harmony_client.submit(request)

Let's see how it's going. This will show the percentage complete in the progress field. (We use the JSON helper function to show the results in a nicer-to-look-at format).

In [5]:

JSON(harmony_client.status(job_id))

Out[5]:

<IPython.core.display.JSON object>

Let's download the results to our system temp directory, overwriting files if they already exist. This returns us a list of Future objects. Each of these "stand in" for a file in our set of results. We can ask a Future for its result and when it's available, it will return the filename to us.

In [6]:

results = harmony_client.download_all(job_id, directory='/tmp', overwrite=True)

Allright, now let's show some colorful Alaska images! Here we iterate over the results, asking each Future for its result, and then using rasterio to open the file and display the image.

In [7]:

for r in results:
    rasterio.plot.show(rasterio.open(r.result()))