Changes

0.7.0 (2026-02-03)

  • Cloud Mask Handling: Resolved a bug in the merge_tifs() function to properly handle “No Data” values in the output cloud mask, ensuring seamless mosaic mask generation.

  • Custom Projections: Added an optional projection_output parameter, allowing users to define a custom CRS for the output mosaic (defaults to EPSG:4326).

  • Automated Grid Cropping: Added an optional grid_crop parameter; when enabled, the final result is automatically clipped to the specific geometry of the input grid tile.

  • Filter_scenes Function: Redesigned the filter_scenes function now use the geometry of grid or bbox to filter the downloaded images for scenes that intersect mosaic region.

  • Mosaic Generation Algorithm Refactor: Completely refactored the merge_scene() function, implementing a new merging logic based on NumPy arrays and binary mask filling to improve mosaic composition.

  • New Example Notebook: Added experimental Jupyter notebooks for function demonstrations:

    • smosaic-introduction: Example of creating RGB mosaics on the BDC_SM_V2 grid for a tile in Mato Grosso state

  • Documentation: Added comprehensive Sphinx documentation with API reference for all functions, usage examples, and detailed parameter descriptions.

  • Command Line Interface: Added a CLI (Command Line Interface) using Click framework, providing command-line access to mosaic generation with configurable options.

0.6.0 (2025-12-22)

  • Standardized Output Resolution: Added new reproject_tif() function that ensures all output mosaics maintain consistent 10,-10 pixel size

  • Projection Consistency: All output mosaics now maintain uniform spatial reference system using the standardized Albers projection

  • Sentinel-2 Baseline Correction Refactor: Redesigned the baseline correction function to properly handle Sentinel-2 processing baseline numbers

  • Enhanced Output Bands Refactor: Completely redesigned mosaic outputs to include provenance and cloud bands by default alongside requested spectral bands

  • Mosaic Profiles: Added profile parameter to mosaic function, providing pre-configured band selections instead of manually selecting bands. Currently supports two profiles:

    • urban_analysis: Pre-selects RGB (Red, Green, Blue) bands along with NDVI and NDBI indices

    • crop_condition: Pre-selects four vegetation indices: NDVI, EVI, EVI2, and SAVI

  • Spectral Indices Calculation: Added smosaic_spectral_indices() function to generate spectral indices mosaics, currently supporting five indices:

    • ndvi: Normalized Difference Vegetation Index

    • evi: Enhanced Vegetation Index

    • evi2: Two-band Enhanced Vegetation Index

    • savi: Soil-Adjusted Vegetation Index

    • ndbi: Normalized Difference Built-up Index

  • New Example Notebooks: Added experimental Jupyter notebooks for article demonstrations:

    • smosaic-introduction: Example of creating RGB mosaics for a scene in Pará state

    • smosaic-bdc-favelas-sp: Experiment on generating image mosaics for monitoring favelas in São Paulo state

    • smosaic-agricultural-monitoring-pr: Experiment on generating image mosaics for agricultural monitoring in Paraná state

  • Sentinel-2/MSI Level-1C: Added full support for S2_L1C_BUNDLE-1 data. 🛰️

  • Fmask external cloud support: Added support for Fmask (algorithm for detecting clouds and cloud shadows).

0.5.0 (2025-11-17)

  • Notebook to Script Migration: Replaced Jupyter notebooks with Python script examples:

    • smosaic-closest-to-date.py: Example of creating mosaics using closest-to-date selection

    • smosaic-data-cube.py: Example of generating temporal data cubes

    • smosaic-mosaic.py: General mosaic generation example

  • New Provenance Band: Added automatic generation of provenance band indicating the origin date of each selected best pixel in the composition

  • Cloud Band Support: Added automatic cloud band generation alongside user-requested bands

  • File Integrity Verification: Added automatic TIFF file corruption check in download_stream() function

  • Sentinel-2 Baseline Correction: Added new function to correct images based on Sentinel-2 processing baseline number

  • Mosaic Algorithm Refactor: Completely redesigned merge_scene() function with improved best-pixel mosaic generation

  • Conda Environment Guide: Added conda-environment.md documentation for creating Conda environments to install and run the library

  • Enhanced Output Bands: All mosaic outputs now include provenance and cloud bands by default alongside requested spectral bands

  • Progress Tracking: Enhanced library logging output with detailed progress tracking, displaying percentage completion per scene and band during mosaic generation

0.4.0 (2025-11-10)

  • New Mosaic Build Methods: Added two new scene ordering functions for mosaic generation:

    • ctd (Closest to Date): Build mosaics by selecting images closest to a reference date

    • chrono (Chronological): Build mosaics by ordering scenes by acquisition date

  • Enhanced Mosaic Function: The mosaic() function now supports three build methods:

    • lcf: Least cloud cover first (existing)

    • ctd: Closest to reference date (new)

    • chrono: Chronological order (new)

  • New Utility Function: Added days_between_dates() function in utils module for date interval calculations

  • New Notebook: Added example notebook:

    • smosaic-closest-to-date.ipynb: A complete example of creating Sentinel-2 image mosaic by selecting images closest to a reference date

0.3.0 (2025-10-31)

  • Architectural Refactor: Split single file smosaic_core.py into modular package structure with specialized modules.

  • Breaking Changes:

    • All functions now distributed across dedicated modules (smosaic_clip_raster, smosaic_merge_scene, etc.)

    • Update imports to reference new modules (e.g., from smosaic_clip_raster import clip_raster)

  • Enhanced Mosaic Function: Added support for monthly periods, with proper date handling.

  • New Notebook: Added example notebook:

    • smosaic-monitoring-expansion-favelas-sp.ipynb: A complete example of creating monthly Sentinel-2 image mosaics for monitoring the expansion of favelas in São Paulo.

0.2.5 (2025-10-18)

  • Fix: Resolved an import error with numpy, pyproj, shapely, requests, rasterio and pystac-client modules.

0.2.2 (2025-10-15)

  • Fix: Fixed a bug in the mosaic function, now it generates both single-date mosaics and data cubes correctly.

0.2.0 (2025-10-10)

  • Multi-band Support: It is now possible to create an mosaic with more than one band.

  • Refactored Library Code: Adjusted imports and the use of libraries in the code, removing imports of individual functions.

  • New Notebooks: Added several example notebooks:

    • smosaic-introduction.ipynb: A complete example of creating a Sentinel-2 multi-band mosaic for Luis Eduardo Magalhaes - BA.

    • smosaic-data-cube.ipynb: A complete example of creating a Sentinel-2 10 days data cube for a given bbox.

  • Data Cube Support: Added support for data cube generation using end_year, end_month, end_day and duration_days parameters.

  • Refactor filter_scenes Function: Completely refactored filter_scenes function now use the grid geometry instead of the colleciton.json file.

  • Implemented parallel processing: to significantly speed up mosaic generation by processing multiple time steps concurrently.✨

0.0.1 (2025-06-04)

  • Initial Release: First implementation of mosaic function, with collection_get_data, get_dataset_extents, merge_tifs and clip_raster functions.

  • Completed the smosaic exemple notebook.

  • Sentinel 2: Added full support for Sentinel 2 data. 🛰️

  • COG Support: Added output as Cloud Optimized GeoTIFFs (COGs) with RasterIO.