TSGenerator 2.0.1 — Maintenance Release

TSGenerator 2.0.1 — Maintenance Release

This maintenance release improves the reliability of Copernicus HR-VPP processing and the usability of the unified Shiny application.

Fixes and improvements

• Fixed VPP scaling for MAXV and TPROD products when GDAL scale/offset metadata have already been applied by terra, preventing double scaling.
• Added regression tests for scaled and unscaled VPP raster inputs.
• Improved Shiny GeoTIFF input validation, temporary-file handling, and direct-disk raster inspection.
• Added a configurable browser upload limit (512 MiB by default, controlled through TSG_SHINY_UPLOAD_MAX_MIB).
• Updated application version display, release metadata, documentation, and citation information.

Compatibility

No intentional changes to the public TSGenerator 2.0 API.

Validation

The release candidate passed 216 local tests, with one intentionally skipped live integration test. Local R CMD check completed with zero errors, warnings, or notes. GitHub CI checks passed across the tested R environments and operating systems.

Citation

A version-specific DOI will be available through Zenodo after the GitHub release has been archived.

TSGenerator 2.0.0

TSGenerator 2.0.0

TSGenerator 2.0.0 is a major release of the R package for acquiring, processing, quality-assessing, and analysing vegetation time series derived from Copernicus High Resolution Vegetation Phenology and Productivity (HR-VPP) products.

Version 2.0 introduces a redesigned architecture centred on native R access to WEkEO, terra-based geospatial processing, explicit product-quality handling, temporal completeness assessment, and an integrated Shiny workflow.

Highlights

• Native R access to the WEkEO Harmonised Data Access service through hdar.
• Acquisition of Copernicus HR-VPP Seasonal Trajectories (ST) and Vegetation Phenology and Productivity (VPP) products.
• terra-based geospatial extraction for polygon-based ST time series and VPP parameters.
• Product-aware quality assessment and masking for current ST/VPP quality information.
• Temporal completeness and missingness diagnostics.
• Optional time-series imputation with safeguards for already processed ST products.
• GAM-based missingness modelling and visualization.
• Unified Shiny application implementing the workflow: Acquire → Prepare/Extract → Quality → Analyze → Export.
• Expanded automated test suite, including live WEkEO integration testing.
• New documentation and vignettes for getting started, ST/VPP workflows, and migration from TSGenerator 1.x.
• Legacy TSGenerator 1.x source retained for traceability and migration support.

Main TSGenerator 2.0 API

Core functions include:

download_st(), download_vpp(), extract_ts(), extract_vpp(), quality_info(), classify_quality(), summarize_quality(), mask_quality(), temporal_plan(), summarize_missingness(), assess_ts_quality(), impute_ts(), model_missingness(), plot_missingness(), check_wekeo(), check_wekeo_integration(), and runTSapp().

Shiny application

The integrated graphical interface can be launched with:

library(TSGenerator)
runTSapp()

The application provides an integrated workflow for data acquisition, spatial preparation and extraction, quality assessment, temporal analysis, and export.

Migration from TSGenerator 1.x

TSGenerator 2.0 replaces several historical workflows with a consolidated API. Selected 1.x functions and historical source code are retained for compatibility and traceability.

Users migrating existing workflows should consult the included migration-1x-to-2x vignette.

Validation

The release was validated through:

• R CMD check: 0 errors, 0 warnings
• Automated package test suite
• Live authenticated WEkEO integration tests: 6 passed, 0 failed, 0 skipped
• Real ST and VPP catalogue queries
• Integrated Shiny workflow tests

The remaining R CMD check note concerning future file timestamps was environment-related and did not indicate a package error.

Scientific use

TSGenerator has been presented and applied in scientific research involving Copernicus HR-VPP products and vegetation time-series processing.

TSGenerator software publication

Valero-Jorge, A., Casterad, M. A., & Alcalá, J.-T. (2026).
TSGenerator: librería R de código abierto para el procesado integrado de productos fenológicos HR-VPP de Copernicus.
Congresos UEx, Actas de Congresos, 2.
DOI: 10.17398/3101-7177.2.123

Research using TSGenerator

Valero-Jorge, A., Casterad, M. A., & Alcalá, J.-T. (2025).
Evaluating the Influence of Missing Data from the Crop Vegetation Index Time Series on Copernicus HR-VPP Phenological Products.
Engineering Proceedings, 94(1), 4.
DOI: 10.3390/engproc2025094004

Authors

• Alexey Valero-Jorge — ORCID: 0000-0002-5993-7346
• Mª Auxiliadora Casterad Seral — ORCID: 0000-0003-4458-6966
• José-Tomás Alcalá Nalvaiz — ORCID: 0000-0001-7549-8825

Funding

TSGenerator was developed within the LAIKcA R&D project (PID2021-124029OR-I00), funded by MICIU/AEI/10.13039/501100011033 and ERDF/EU.

Alexey Valero-Jorge acknowledges predoctoral grant PRE2022-102328, funded by MICIU/AEI/10.13039/501100011033 and ESF+.

License

TSGenerator 2.0.0 is distributed under the MIT License.

DOI

TSGenerator is archived in Zenodo.

A version-specific DOI for TSGenerator 2.0.0 will be assigned after archival of this release.

vv1.0.0.2025

The TSGenerator package has been created with the aim of facilitating time series extraction of HRVPP data from Copernicus for users with different levels of R programming.