stackstac

https://stackstac.readthedocs.io/en/latest/
Admin
stackstac is a Python library that converts STAC collections into lazy xarray DataArrays backed by...
Tokens:26,136
Snippets:144
Trust Score:8.3
Update:7 months ago
Show doc for...
Tokens:
Raw
### Install stackstac library

Source: https://stackstac.readthedocs.io/en/latest/index.html

Installs the stackstac package along with optional visualization dependencies using pip. This command ensures all necessary components for cloud-native geoprocessing are available in the environment.

```bash
pip install 'stackstac[viz]'
```

--------------------------------

### Configure Stackstac parameters

Source: https://stackstac.readthedocs.io/en/latest/api/main/stackstac.stack.html

Example configuration demonstrating how to set spatial bounds, chunking, and data types for a Stackstac operation. These parameters control how STAC items are loaded into a Dask-backed xarray DataArray.

```python
import numpy as np
from rasterio.enums import Resampling

# Example configuration for stackstac.stack
config = {
    "bounds": (10.0, 45.0, 12.0, 47.0),  # (min_x, min_y, max_x, max_y)
    "snap_bounds": True,
    "resampling": Resampling.nearest,
    "chunksize": (1, 1, 512, 512),
    "dtype": "float32",
    "fill_value": np.nan,
    "rescale": True,
    "sortby_date": "asc",
    "xy_coords": "topleft"
}
```

--------------------------------

### PandasIndex Examples

Source: https://stackstac.readthedocs.io/en/latest/basic.html

Demonstrates the creation and structure of PandasIndex objects for different data types (object, float64, datetime64). These indices are commonly used in libraries like xarray to label data dimensions.

```python
PandasIndex(Index(['red', 'green', 'blue'], dtype='object', name='band'))
```

```python
PandasIndex(Float64Index([399960.0, 399970.0, 399980.0, 399990.0, 400000.0, 400010.0,
              400020.0, 400030.0, 400040.0, 400050.0,
              ... 
              509660.0, 509670.0, 509680.0, 509690.0, 509700.0, 509710.0,
              509720.0, 509730.0, 509740.0, 509750.0],
             dtype='float64', name='x', length=10980))
```

```python
PandasIndex(Float64Index([4000020.0, 4000010.0, 4000000.0, 3999990.0, 3999980.0, 3999970.0,
              3999960.0, 3999950.0, 3999940.0, 3999930.0,
              ... 
              3890320.0, 3890310.0, 3890300.0, 3890290.0, 3890280.0, 3890270.0,
              3890260.0, 3890250.0, 3890240.0, 3890230.0],
             dtype='float64', name='y', length=10980))
```

```python
PandasIndex(DatetimeIndex(['2020-03-01', '2020-04-01', '2020-05-01'], dtype='datetime64[ns]', name='time', freq='MS'))
```

--------------------------------

### stackstac.rio_env.LayeredEnv

Source: https://stackstac.readthedocs.io/en/latest/api/internal/stackstac.rio_env.LayeredEnv.html

Manages GDAL configuration options for different situations (always, opening, reading). Access the rasterio.Env for each situation with layered_env.always, layered_env.read, etc. To ensure thread-safety, each thread accessing these properties gets its own thread-local copy of a rasterio.Env.

```APIDOC
## stackstac.rio_env.LayeredEnv

### Description
Manages GDAL configuration options for different situations (always, opening, reading). Access the `rasterio.Env` for each situation with `layered_env.always`, `layered_env.read`, etc. To ensure thread-safety, each thread accessing these properties gets its own thread-local copy of a `rasterio.Env`.

Options are layered as follows:
  * `always`: the base set of options
  * `open`: `always` + `open`
  * `open_vrt`: `open` + `open_vrt`
  * `read`: `always` + `read`

### Methods

#### `__init__`

```python
__init__(_always_, _open=None_, _open_vrt=None_, _read=None_)
```

Initializes the LayeredEnv with base options and optional overrides for opening, opening VRTs, and reading.

##### Parameters

- **_always** (rasterio.Env) - Required - The base set of GDAL options.
- **_open** (rasterio.Env, optional) - The GDAL options to apply when opening datasets.
- **_open_vrt** (rasterio.Env, optional) - The GDAL options to apply when opening VRT datasets.
- **_read** (rasterio.Env, optional) - The GDAL options to apply when reading from datasets.

#### `updated`

```python
updated([always, open, open_vrt, read])
```

Duplicates this LayeredEnv, adding additional options for each situation.

##### Parameters

- **always** (rasterio.Env, optional) - Additional base GDAL options.
- **open** (rasterio.Env, optional) - Additional GDAL options for opening datasets.
- **open_vrt** (rasterio.Env, optional) - Additional GDAL options for opening VRT datasets.
- **read** (rasterio.Env, optional) - Additional GDAL options for reading from datasets.

### Attributes

- **`always`** (rasterio.Env) - The base `rasterio.Env` object.
- **`open`** (rasterio.Env) - The `rasterio.Env` object to use while opening datasets.
- **`open_vrt`** (rasterio.Env) - The `rasterio.Env` object to use while opening VRTs.
- **`read`** (rasterio.Env) - The `rasterio.Env` object to use while reading from datasets.

### Example Usage

```python
import rasterio
from stackstac.rio_env import LayeredEnv

# Define base options
base_env = rasterio.Env(GDAL_TIFF_OVR_BLOCKSIZE=128)

# Create a LayeredEnv
layered_env = LayeredEnv(always=base_env, read=rasterio.Env(GDAL_CACHEMAX=0))

# Access different environments
with layered_env.read as env:
    # Use env for reading operations
    print(env.get('GDAL_CACHEMAX')) # Output: 0

with layered_env.always as env:
    # Use env for base operations
    print(env.get('GDAL_TIFF_OVR_BLOCKSIZE')) # Output: 128
```
```

--------------------------------

### Accessing Band Metadata and Spatial Indexes

Source: https://stackstac.readthedocs.io/en/latest/examples/cluster.html

Examples of how band-specific attributes like common names, wavelengths, and spatial resolutions are stored, alongside the definition of temporal and spatial indexes (x, y, time) within the data structure.

```python
import numpy as np
import pandas as pd

# Example of band metadata extraction
common_names = np.array([None, 'blue', 'coastal', 'green', 'nir', 'nir08', 'nir09', 'red', 'rededge', 'rededge', 'rededge', None, 'swir16', 'swir22', None, None], dtype=object)

# Example of spatial index definition
x_index = pd.Index([391300.0, 391400.0, 391500.0], name='x')
y_index = pd.Index([3984700.0, 3984600.0, 3984500.0], name='y')

# EPSG code representation
epsg = 32613
```

--------------------------------

### Initialize Coiled Dask Cluster

Source: https://stackstac.readthedocs.io/en/latest/examples/gif.html

Configures and connects to a Coiled Dask cluster to enable parallel processing of large geospatial datasets. This setup optimizes performance by running computations in a region close to the data source.

```python
cluster = coiled.Cluster(
    name="stackstac-eu",
    n_workers=20,
    package_sync=True,
    backend_options={"region": "eu-central-1"}
)
client = distributed.Client(cluster)
client
```

--------------------------------

### Reproject and Set Resolution using stackstac.stack

Source: https://stackstac.readthedocs.io/en/latest/api/main/stackstac.stack.html

This example shows how to reproject the STAC items to a specified Coordinate Reference System (CRS) and resolution using the `epsg` and `resolution` parameters. It's crucial that the resolution is provided in the units of the output CRS.

```python
import stackstac
import satsearch

items = satsearch.Search(...).items()

# Reproject to 100-meter resolution in web mercator
xr_stack = stackstac.stack(items, epsg=3857, resolution=100)
```

--------------------------------

### Compute Dask Array with Progress Bar

Source: https://stackstac.readthedocs.io/en/latest/basic.html

Demonstrates how to trigger the computation of a lazy Dask array using a context manager to display a progress bar. This is essential for monitoring long-running geospatial data processing tasks.

```python
import dask.diagnostics
with dask.diagnostics.ProgressBar():
    data = aoi.compute()
```

--------------------------------

### Filtering Satellite Data by Cloud Cover

Source: https://stackstac.readthedocs.io/en/latest/examples/gif.html

This example shows how to filter satellite data based on cloud cover using the 'eo:cloud_cover' attribute. It assumes the existence of a data structure (like a GeoDataFrame or xarray DataArray) with this metadata, allowing for the selection of scenes with minimal cloud presence.

```python
import xarray as xr
import numpy as np

# Assume 'data' is an xarray DataArray with satellite imagery and metadata
# For demonstration, we create a dummy DataArray with a 'cloud_cover' attribute

data = xr.DataArray(
    np.random.rand(10, 10),
    dims=('y', 'x'),
    coords={'y': np.arange(10), 'x': np.arange(10)},
    attrs={'eo:cloud_cover': np.random.rand(10) * 100} # Dummy cloud cover values
)

# Define the maximum acceptable cloud cover percentage
max_cloud_cover = 20

# Filter the data to include only scenes with cloud cover less than the threshold
filtered_data = data.where(data.attrs['eo:cloud_cover'] < max_cloud_cover, drop=True)

print(f"Original data shape: {data.shape}")
print(f"Filtered data shape (cloud cover < {max_cloud_cover}%): {filtered_data.shape}")
print(f"Cloud cover values in filtered data: {filtered_data.attrs['eo:cloud_cover']}")

```

--------------------------------

### Initialize STAC Catalog and Search

Source: https://stackstac.readthedocs.io/en/latest/basic.html

Connects to the Earth Search AWS STAC catalog and performs a spatial and temporal query for Sentinel-2 data. Returns an item collection containing metadata for the requested scenes.

```python
import pystac_client
import stackstac

lon, lat = -105.78, 35.79
URL = "https://earth-search.aws.element84.com/v1"
catalog = pystac_client.Client.open(URL)

items = catalog.search(
    intersects=dict(type="Point", coordinates=[lon, lat]),
    collections=["sentinel-2-l2a"],
    datetime="2020-03-01/2020-06-01"
).item_collection()
```

--------------------------------

### Custom Dtype and Fill Value with stackstac.stack

Source: https://stackstac.readthedocs.io/en/latest/api/main/stackstac.stack.html

This example demonstrates how to specify a custom data type (`dtype`) and fill value (`fill_value`) for the output DataArray. The `rescale` parameter is also set to `False`, indicating that no automatic rescaling of data values should be performed.

```python
import stackstac
import satsearch

items = satsearch.Search(...).items()

# Custom dtype and fill_value
xr_stack = stackstac.stack(items, rescale=False, fill_value=0, dtype="uint16")
```

--------------------------------

### Initialize Interactive Map with ipyleaflet

Source: https://stackstac.readthedocs.io/en/latest/examples/show.html

Sets up an interactive map centered on a specific coordinate with a defined zoom level and height for visualization.

```python
m = ipyleaflet.Map()
m.center = 35.677153763176115, -105.8485489524901
m.zoom = 10
m.layout.height = "700px"
m
```

--------------------------------

### Spatial Subsetting with xarray .loc

Source: https://stackstac.readthedocs.io/en/latest/basic.html

Selects a spatial subset of an xarray DataArray using coordinate ranges. This example demonstrates slicing based on UTM coordinates and a buffer, effectively extracting a region of interest.

```python
aoi = monthly.loc[..., y_utm+buffer:y_utm-buffer, x_utm-buffer:x_utm+buffer]
aoi
```

--------------------------------

### Import required geospatial and distributed computing libraries

Source: https://stackstac.readthedocs.io/en/latest/examples/show.html

Initializes the environment by importing necessary modules for stackstac, STAC client interaction, map visualization, and Dask distributed computing.

```python
import stackstac
import pystac_client
import ipyleaflet
import xarray as xr
import IPython.display as dsp
import coiled
import distributed
```

--------------------------------

### Function: stackstac.show

Source: https://stackstac.readthedocs.io/en/latest/api/main/stackstac.show.html

The show function creates an ipyleaflet.Map to visualize DataArrays. It computes data on the fly as the user pans, requiring a dask distributed cluster.

```APIDOC
## FUNCTION stackstac.show

### Description
Quickly create an `ipyleaflet.Map` displaying a `DataArray`. As you pan around the map, the part of the array that is in view is computed on the fly by dask.

### Method
Python Function

### Parameters
#### Arguments
- **arr** (`DataArray`) - Required - The DataArray to visualize. Must have x and y dimensions and an epsg coordinate set.
- **center** (`Tuple[float, float]`) - Optional - Centerpoint for the map in (lat, lon) order. Defaults to auto-centering.
- **zoom** (`int`) - Optional - Initial zoom level for the map.
- **range** (`Tuple[float, float]`) - Optional - Min and max values for visualization scaling. Defaults to 2nd/98th percentile.
- **cmap** (`Union[str, Colormap]`) - Optional - Colormap for single-band data. Defaults to 'viridis'.
- **checkerboard** (`bool`) - Optional - Whether to show a checkerboard pattern for missing data. Defaults to True.
- **interpolation** (`Literal['linear', 'nearest']`) - Optional - Interpolation method for reprojection. Defaults to 'linear'.

### Request Example
```python
import stackstac
stackstac.show(my_data_array, cmap='magma', interpolation='nearest')
```

### Response
#### Success Response
- **return** (`ipyleaflet.Map`) - The interactive map object displaying the array.
```

--------------------------------

### Coordinate Transformation with pyproj

Source: https://stackstac.readthedocs.io/en/latest/basic.html

Converts longitude and latitude coordinates to UTM coordinates using the pyproj library. This is a crucial step for aligning geographic points with projected raster data.

```python
import pyproj
x_utm, y_utm = pyproj.Proj(monthly.crs)(lon, lat)
buffer = 2000  # meters
```

--------------------------------

### Create and Connect to a Coiled Dask Cluster

Source: https://stackstac.readthedocs.io/en/latest/examples/cluster.html

This snippet demonstrates how to initialize a Dask cluster using Coiled, specifying the number of workers, package synchronization, and the cloud region for backend operations. It then establishes a client connection to this cluster.

```python
import coiled
import distributed

cluster = coiled.Cluster(
    name="stackstac",
    n_workers=8,
    package_sync=True,
    backend_options={"region": "us-west-2"},  # where the data is
)
client = distributed.Client(cluster)
client
```

--------------------------------

### Execute Parallel Computation and Visualization

Source: https://stackstac.readthedocs.io/en/latest/examples/cluster.html

Demonstrates initializing a compute cluster, executing a parallel computation task, and plotting the resulting multi-dimensional imagery.

```python
client.wait_for_workers(8)
%time rgb_ = monthly_rgb.compute()
rgb_.plot.imshow(row="time", rgb="band", robust=True, size=6);
```

--------------------------------

### xarray DataArray Representation

Source: https://stackstac.readthedocs.io/en/latest/basic.html

Shows the detailed representation of an xarray DataArray after spatial subsetting. It includes dimensions, dask array information, coordinates, and attributes, providing insights into the data structure and its metadata.

```python
xarray.DataArray
'stackstac-3a21164f19f81366a5242c365d798d31'
  * time: 3
  * band: 3
  * y: 400
  * x: 400


  * dask.array<chunksize=(1, 1, 387, 316), meta=np.ndarray>
|  |  Array  |  Chunk   
---|---|---
Bytes  |  10.99 MiB  |  0.93 MiB   
Shape  |  (3, 3, 400, 400)  |  (1, 1, 387, 316)   
Dask graph  |  36 chunks in 16 graph layers   
Data type  |  float64 numpy.ndarray   
  * Coordinates: (21)
    * band
(band)
<U12
'red' 'green' 'blue'
```
array(['red', 'green', 'blue'], dtype='<U12')
```

    * x
(x)
float64
4.275e+05 4.275e+05 ... 4.315e+05
```
array([427520., 427530., 427540., ..., 431490., 431500., 431510.])
```

    * y
(y)
float64
3.963e+06 3.963e+06 ... 3.959e+06
```
array([3962930., 3962920., 3962910., ..., 3958960., 3958950., 3958940.])
```

    * s2:product_type
()
<U7
'S2MSI2A'
```
array('S2MSI2A', dtype='<U7')
```

    * constellation
()
<U10
'sentinel-2'
```
array('sentinel-2', dtype='<U10')
```

    * mgrs:grid_square
()
<U2
'DV'
```
array('DV', dtype='<U2')
```

    * grid:code
()
<U10
'MGRS-13SDV'
```
array('MGRS-13SDV', dtype='<U10')
```

    * proj:epsg
()
int64
32613
```
array(32613)
```

    * instruments
()
<U3
'msi'
```
array('msi', dtype='<U3')
```

    * mgrs:latitude_band
()
<U1
'S'
```
array('S', dtype='<U1')
```

    * s2:saturated_defective_pixel_percentage
()
int64
0
```
array(0)
```

    * mgrs:utm_zone
()
int64
13
```
array(13)
```

    * s2:datatake_type
()
<U8
'INS-NOBS'
```
array('INS-NOBS', dtype='<U8')
```

    * title
(band)
<U31
'Red (band 4) - 10m' ... 'Blue (...
```
array(['Red (band 4) - 10m', 'Green (band 3) - 10m',
       'Blue (band 2) - 10m'], dtype='<U31')
```

    * raster:bands
(band)
object
None None None
```
```

--------------------------------

### Import Geospatial and Data Processing Libraries

Source: https://stackstac.readthedocs.io/en/latest/examples/gif.html

Imports the essential Python libraries required for satellite data retrieval, distributed computing, and visualization. Dependencies include stackstac, pystac_client, planetary_computer, and dask.

```python
import coiled
import distributed
import dask
import pystac_client
import planetary_computer as pc
import ipyleaflet
import IPython.display as dsp
import geogif
import stackstac
```

--------------------------------

### Clip to Custom Bounding Box with stackstac.stack

Source: https://stackstac.readthedocs.io/en/latest/api/main/stackstac.stack.html

This example demonstrates clipping the output DataArray to a custom bounding box defined by latitude and longitude coordinates using the `bounds_latlon` parameter. This allows focusing on a specific geographic region.

```python
import stackstac
import satsearch

items = satsearch.Search(...).items()

# Clip to a custom bounding box
xr_stack = stackstac.stack(items, bounds_latlon=[-106.2, 35.6, -105.6, 36])
```

--------------------------------

### Initialize Interactive Map with ipyleaflet

Source: https://stackstac.readthedocs.io/en/latest/examples/gif.html

Configures an interactive map interface using ipyleaflet to select a specific geographic area. The map is set with a default center, zoom level, and height for user interaction.

```python
m = ipyleaflet.Map(scroll_wheel_zoom=True)
m.center = 41.64933994767867, -69.94438630063088
m.zoom = 12
m.layout.height = "800px"
m
```

--------------------------------

### Visualize Geospatial Imagery

Source: https://stackstac.readthedocs.io/en/latest/basic.html

Uses xarray's plotting capabilities to render multi-dimensional imagery. It maps the 'time' dimension to rows and 'band' dimension to RGB channels, applying robust scaling for visualization.

```python
data.plot.imshow(row="time", rgb="band", robust=True, size=6);
```

--------------------------------

### Inspect StackSTAC Data Arrays

Source: https://stackstac.readthedocs.io/en/latest/examples/gif.html

Demonstrates the structure of various data arrays in StackSTAC, including time-series float data and categorical metadata like instrument types and Landsat collection parameters.

```python
# Example of time-series float64 data
import numpy as np
data = np.array([9.294e+01, 7.700e-01, 1.181e+01])

# Instrument metadata
instruments = np.array({'tirs', 'oli'}, dtype=object)

# Landsat processing level
proc_level = np.array('L2SP', dtype='<U4')

# Solar elevation data
sun_elevation = np.array([44.94417071, 50.03688029, 51.91632915])
```

--------------------------------

### Stack STAC Items into Xarray

Source: https://stackstac.readthedocs.io/en/latest/basic.html

Converts a collection of STAC items into a lazy xarray DataArray. This process uses dask to handle large datasets by creating chunks, allowing for efficient memory management.

```python
stack = stackstac.stack(items)
```

--------------------------------

### Reader Class Documentation

Source: https://stackstac.readthedocs.io/en/latest/api/internal/stackstac.reader_protocol.Reader.html

Documentation for the Reader class, including its constructor and methods.

```APIDOC
## Class stackstac.reader_protocol.Reader

### Description
Protocol for a thread-safe, lazily-loaded object for reading data from a single-band STAC asset.

### Methods

#### `__init__(*args, **kwargs)`

##### Description
Initializes the Reader object.

##### Method
`__init__`

##### Parameters
* `*args`: Variable length argument list.
* `**kwargs`: Arbitrary keyword arguments.

#### `close()`

##### Description
Close all resources for the Dataset.

##### Method
`close`

#### `read(window)`

##### Description
Read a portion of data from the Dataset as a NumPy array.

##### Method
`read`

##### Parameters
* `window` (object): The window to read from the dataset.

##### Returns
* NumPy array: The data read from the specified window.
```

--------------------------------

### Interactively Explore NDVI with ipywidgets and stackstac

Source: https://stackstac.readthedocs.io/en/latest/examples/show.html

Uses ipywidgets.interact to create a slider for interactively thresholding NDVI values. This demonstrates integrating stackstac with the Python ecosystem for powerful data exploration, although it can be slow to refresh.

```python
import ipywidgets

ndvi_map = ipyleaflet.Map()
ndvi_map.center = m.center
ndvi_map.zoom = m.zoom

@ipywidgets.interact(threshold=(0.0, 1.0, 0.1))
def explore_ndvi(threshold=0.2):
    high_ndvi = ndvi_comp.where(ndvi_comp > threshold)
    stackstac.add_to_map(high_ndvi, ndvi_map, "ndvi", range=[0, 1], cmap="YlGn")
    return ndvi_map
```

--------------------------------

### Filter and Resample Satellite Data

Source: https://stackstac.readthedocs.io/en/latest/basic.html

Filters a stack of satellite imagery by cloud coverage, selects RGB bands, and performs a monthly median resampling. This process uses xarray's indexing and resampling capabilities to produce aggregated temporal composites.

```python
lowcloud = stack[stack["eo:cloud_cover"] < 20]
rgb = lowcloud.sel(band=["red", "green", "blue"])
monthly = rgb.resample(time="MS").median("time", keep_attrs=True)
```

--------------------------------

### Configure and connect to a Coiled Dask cluster

Source: https://stackstac.readthedocs.io/en/latest/examples/show.html

Sets up a remote Dask cluster using Coiled to perform heavy data computation near the data source. The cluster is configured with specific worker resources and region settings for optimal performance.

```python
cluster = coiled.Cluster(
    name="stackstac-show",
    n_workers=22,
    worker_cpu=4,
    worker_memory="16GiB",
    package_sync=True,
    backend_options={"region": "us-west-2"},
)
client = distributed.Client(cluster)
client
```

--------------------------------

### Analyze temporal intervals

Source: https://stackstac.readthedocs.io/en/latest/examples/gif.html

Calculates and visualizes the temporal distribution of satellite scenes by computing the difference between consecutive time indices.

```python
# What's the typical interval between scenes?
good.time.diff("time").dt.days.plot.hist();
```

--------------------------------

### SingleThreadedRioDataset Class Initialization (Python)

Source: https://stackstac.readthedocs.io/en/latest/api/internal/stackstac.rio_reader.SingleThreadedRioDataset.html

Initializes the SingleThreadedRioDataset, an interface for rasterio datasets with single-threaded drivers. It requires an environment, a dataset, and optionally a VRT. Concurrent reads are protected by a lock.

```python
class SingleThreadedRioDataset:
    """
    Interface for a rasterio dataset whose driver is inherently single-threaded (like hdf5).
    Concurrent reads are protected by a lock. 
    """
    def __init__(self, env, ds, vrt=None):
        """
        Initializes the SingleThreadedRioDataset.

        Args:
            env: The rasterio environment.
            ds: The rasterio dataset.
            vrt: Optional VRT dataset.
        """
        pass
```

--------------------------------

### Load Landsat WRS Paths with Stackstac

Source: https://stackstac.readthedocs.io/en/latest/examples/gif.html

This snippet shows how to extract the World Reference System (WRS) path for Landsat scenes using Stackstac. It assumes the data is loaded and the 'wrs_path' attribute is available. The output is a NumPy array of strings representing the WRS paths.

```python
import numpy as np

# Assuming 'data' is a loaded dataset object with a 'wrs_path' attribute
# Example placeholder for demonstration:
data = {
    'wrs_path': np.array(['011', '012', '012', '012', '012', '011', '012', '011', '012', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '011', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '012', '011', '012', '011', '012', '011', '012', '011', '011', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '011', '012', '
```

--------------------------------

### Query STAC Catalog for Sentinel-2 Data

Source: https://stackstac.readthedocs.io/en/latest/examples/show.html

Uses pystac_client to search the Earth Search AWS STAC API for Sentinel-2 L2A collections within a defined bounding box and date range.

```python
bbox=[m.west, m.south, m.east, m.north]
stac_items = pystac_client.Client.open(
    "https://earth-search.aws.element84.com/v1"
).search(
    bbox=bbox,
    collections=["sentinel-2-l2a"],
    datetime="2020-04-01/2020-04-15"
).item_collection()
```

--------------------------------

### Create monthly median RGB composites

Source: https://stackstac.readthedocs.io/en/latest/examples/cluster.html

Uses xarray's selection and resampling methods to isolate RGB bands and compute the median value for each month in the dataset.

```python
rgb = stack.sel(band=["red", "green", "blue"])
monthly_rgb = rgb.resample(time="MS").median(dim="time")
monthly_rgb
```

--------------------------------

### Create Time Stack with stackstac

Source: https://stackstac.readthedocs.io/en/latest/examples/show.html

Converts a collection of STAC items into a stackstac data cube with a specified spatial resolution.

```python
stack = stackstac.stack(stac_items, resolution=80)
```
Tokens:
Raw