Coverage analysis tool for optical data
=======================================
*Check per-pixel data availability*

This module uses the `sepal_ui <https://github.com/12rambau/sepal_ui>`_ framework and interactive **Voila** dashboard to create maps of cloud-free observations for major optical satellites available on the Google Earth Engine (GEE) platform.

The framework follows the logic of BFAST's countObs and summaryBrick functions, as described `here <http://www.loicdutrieux.net/bfastSpatial/#Data_Inventory>`_ (BFAST refers to Breaks For Additive Season and Trend). 

For more information about BFAST, see `Schultz et al. (2013) <http://dx.doi.org/10.1109/JSTARS.2015.2477473>`_.

The three-step process is described in the following sections.

Select an area of interest (AOI)
--------------------------------

Using the provided **AOI** selector, select an AOI from the available methods. We provide three administrative descriptors (from level 0 to 2) and three custom shapes (drawn directly on the map, asset or shapefile).

.. figure:: https://raw.githubusercontent.com/BuddyVolly/coverage_analysis/main/doc/img/aoi_select.png 
    
    AOI selector 
    
.. note::

    If a custom AOI from a shape or drawing is selected, you will be able to use it directly (the upload to GEE will be launched in the background).
    
Select dataset parameters
-------------------------

To perform BFAST pre-analysis, provide the tool with key parameters:

-   **Date range**: the start and end dates of your analysis
-   **Sensors**: the list of sensors you want to use (Landsat missions and Sentinel-2)
-   **Tier 2**: Tier 2 images of the Landsat missions (note: this might lead to incorrect results)
-   **SR**: whether to use surface reflectance (SR) images (by default, TOA, referring to top of atmosphere)

Once all parameters have been chosen, select the button.

.. figure:: https://raw.githubusercontent.com/BuddyVolly/coverage_analysis/main/doc/img/parameters.png 

Display dataset
---------------

After selecting your parameters, move to the **Visualization** tile.

Select one of the statistical measures to display in the following list:

-   cloud-free pixel count
-   total pixel count (i.e. scene coverage)
-   NDVI Median (normalized difference vegetation index median)
-   NDVI Std. Dev. (normalized difference vegetation index standard deviation)

You can also produce stats on a yearly basis using the provided switch.

.. figure:: https://raw.githubusercontent.com/BuddyVolly/coverage_analysis/main/doc/img/display.png 

.. note::

    The image will be dynamically re-evaluated and recentred upon each change.

Export dataset
--------------

When you're satisifed with the displayed information, it can be exported for further use in GIS software or a GEE process. The tool provides two main exportation options: 

-    as an asset (in GEE), or 
-    a .tif file (in SEPAL).

Both use the GEE export system and share the same set of parameters:

-   **Statistical measures to export**
    
    -   count of cloud-free observation per pixel
    -   NDVI's median of cloud-free observations
    -   NDVI's std. dev. of cloud-free observations
    -   count for all observations per pixel

-   **Time period**
    
    -   full timespan calculation(s)
    -   annual calculation(s)

-   **Scale**: the resolution (in metres) to use in the exported GEE file

.. figure:: https://raw.githubusercontent.com/BuddyVolly/coverage_analysis/main/doc/img/export.png 

.. attention::

    When exporting the image to SEPAL, you cannot quit the application until the download is finished.

.. custom-edit:: https://raw.githubusercontent.com/sepal-contrib/coverage_analysis/release/doc/en.rst