Why SISPPEO ?¶
SISPPEO allows scientists to write remote sensing algorithms without having to worry about which satellite product will be used - specific steps like reading/extracting data, or formatting output products are already implemented.
Moreover, this package includes various carefully selected algorithms from the literature, allowing users to derive for example Chl-a concentrations or water turbidity in a river, or to detect burnt areas (NBR).
How does it work ?¶
Data is extracted using “readers” (one per (satellite / atmospheric correction) couple).
The chosen algorithm is applied.
An output product (called a “L3 product”) is built and saved as a well-formatted NetCDF file (CF-1.8).
How to add an algorithm to SISPPEO ?¶
Basically, you just need to write a function that takes n bands
(as xr.DataArray) as parameters and returns one or more arrays.
Which bands are used (e.g., depending the satellite) is defined in a YAML file:
ndwi:
long_name: Normalized Difference Water Index
output: ndwi
L8:
- B3
- B5
S2:
- B3
- B8
For more details, please see the corresponding tutorial.
What if there is no reader for my product ?¶
For now, SISPPEO can read:
S2 products from ESA (both L1C and L2A)
L8 products from USGS (both L1C1 and L2C1)
S2 L2A and L8 L2 GRS products (T. Harmel et al., 2018)
S2 L2A C2RCC products (Brockmann Consult)
S2 L2A MAJA products (O. Hagolle et al., 2017)
You can write a new reader and add it to SISPPEO. It is really super easy to do (cf. this tutorial).