pySICE

Python scripts behind the SICE toolchain for albedo retrieval.

by B. Vandecrux (2), A. Kokhanovsky (1), J. Box (2)

(1) German Research Centre for Geosciences, Potsdam, Germany (2) Geological Survey of Denmark and Greenland (GEUS) Øster Voldgade 10, 1350 Copenhagen, Denmark

Table of Contents

• Running environment
• Theoretical background
• Script overview
• Input preparation
• Outputs
• Installation
• Example data
• Python interface for the fortran script sice.f

Running environment

developped on Python 3.7.6 uses numpy, rasterio, time and sys packages

Theoretical background

The snow surface characteristics retrieval is based on the following work:

• Kokhanovsky et al. (2018) On the reflectance spectroscopy of snow
• Kokhanovsky et al. (2019) Retrieval of Snow Properties from the Sentinel-3 Ocean and Land Colour Instrument
• Kokhanovsky et al. (2020) The Determination of Snow Albedo from Satellite Measurements Using Fast Atmospheric Correction Technique
• Vandecrux et al. (2022) The Determination of the Snow Optical Grain Diameter and Snowmelt Area on the Greenland Ice Sheet Using Spaceborne Optical Observations

The ozone total ozone retreival is described in

• Kokhanovsky et al. (2020) Retrieval of the total ozone over Antarctica using Sentinel-3 ocean and land colour instrument

The Algorithm Theoretical Basis Document is available here

Scripts overview

Input needed

Input files:

File	Description
height.tif	Height in metre
mask.tif	Ice mask
O3.tif	Total column ozone
OZA.tif	Observation zenith angle
OAA.tif	Observation azimuth angle
r_TOA_01..21.tif	Top of the atmosphere OLCI reflectance
SZA.tif	Solar Zenith angle
SAA.tif	Solar azimuth angle

Outputs

File Name	Description	Units
grain_diameter.tif	Snow grain diameter	mm
snow_specific_surface_area.tif	Snow specific surface area	m^{2}kg^{-1}
albedo_bb_planar_sw.tif	shortwave broadband planar albedo	300-2400 nm
albedo_bb_spherical_sw.tif	shortwave broadband spherical albedo	300-2400 nm
rBRR_01..21.tif	bottom of the atmosphere surface reflectance (OLCI bands 1 through 21)
albedo_spectral_planar_01..21.tif	spectral planar albedo (OLCI bands 1 to 11 and 16 to 21)
albedo_spectral_spherical_01..21.tif	spectral spherical albedo (OLCI bands 1 to 11 and 16 to 21)
isnow.tif	See next table
impurity_load.tif	pollutant load	ppm_weight
pol_type.tif	type of pollutant: 1(soot), 2( dust), 3 and 4 (other or mixture)
al.tif	effective absorption length	mm
r0.tif	reflectance of a semi-infinite non-absorbing snow layer
O3_SICE.tif	total ozone product (OLCI) corrected for ozone scattering
factor.tif	Snow covered fraction within mixed pixels
cv1.tif	quality check 1 (see ATBD)
cv2.tif	quality check 1 (see ATBD)

Diagnostic Code	Description
1	clean snow
2	polluted snow
3	partially snow covered pixel
100	sza<75, no retrival
102	TOA reflectance at band 21 < 0.1, no retrieval
103	TOA reflectance at band 1 < 0.2, no retrieval
104	grain_diameter < 0.1, no retrieval, potential cloud flag
105	retrieved spherical albedo negative in band 1, 2 or 3
-n	impossible to solve snow spherical albedo equation at band n

Installation

We recommend the use of Anaconda and recent version of Python (>3.7).

The package can be installed as using

pip install git+https://github.com/GEUS-SICE/[email protected]                                  

The package can be run using

pysice InputFolder OutputFolder

The output is added to the OutputFolder if it exists, if not is is added to the InputFolder folder. See help for other possibilities:

pysice -h


usage: pysice.py [-h] [-i FL_IN] [-o FL_OUT] [-c [CLEAN_SNOW]]
                 [--no_qc [NO_QC]] [--no_oz [NO_OZ]]
                 [input_folder] [output_folder]

positional arguments:
  input_folder          Path to input folder, containing OLCI radiance files
  output_folder         Path to output folder (same as input folder if non-
                        existent or not specified)

optional arguments:
  -h, --help            show this help message and exit
  -i FL_IN, --fl_in FL_IN, --input FL_IN
                        Path to input folder, containing OLCI radiance files
                        (or use first positional argument)
  -o FL_OUT, --fl_out FL_OUT, --output FL_OUT
                        Path to output folder (or use last positional
                        argument)
  -c [CLEAN_SNOW], --clean_snow [CLEAN_SNOW]
                        If present, processes all pixels as clean snow
  --no_qc [NO_QC]       If present, does not run quality check
  --no_oz [NO_OZ]       If present, does not retrieve ozone


Post questions, suggestions, patches at https://github.com/GEUS-SICE/pySICE

Example data

Test input files are available here. Download and unzip using browser or with:

wget https://www.dropbox.com/s/b7wbervqls0p5cc/S3_test_data.zip
unzip S3_test_data.zip -d S3_test_data

Python interface for the fortran script sice.f

sice_f.py reads the SICE-generated S3 geotiff files, converts them into ascii files, compiles and runs sice.f, reads the text outputs and save them into geotiff again.

Compile sice.f:

gfortran ./fortran/sice.f -o ./fortran/sice.exe

Create the output folder and run the script:

mkdir S3_test_data/fortran
python fortran/sice_f.py ./S3_test_data