Advancing Land Monitoring Through Synergistic Harmonization of Optical, Radar and Lidar Satellite Technologies: 2nd Edition

Dear Colleagues,

Land monitoring is the systematic, repeatable observation, measurement, quality control, and analysis of the Earth’s terrestrial surface—its biophysical state, human use, and change—so that decision-ready indicators with known uncertainty and lineage can be produced for management, compliance, and risk reduction. It spans end-to-end workflows: acquiring multi-sensor observations (optical, SAR, lidar), calibrating and co-registering them, harmonizing across sensors and time, modeling and inferring variables, validating against independent data, quantifying uncertainty, and delivering reproducible, operational maps, time series, alerts, and reports. Variables and themes include land cover and land use; vegetation phenology, structure, and biomass; soil moisture, condition, and degradation; surface water, wetlands, and peatlands; terrain and geomorphology; agriculture (crop type, growth stage, yield); the urban and built environment (expansion, density, heat, green space); geohazards (fires, drought, floods, landslides, subsidence); and biodiversity proxies and habitat condition.

A new wave of global, high-cadence, multi-sensor missions is generating unprecedented volumes of complementary data, while cloud platforms and modern ML make routine cross-sensor products feasible; simultaneously, climate action, nature accounting, and disaster-risk mandates demand consistent, defensible land indicators traceable across sensors and time; coupled with growing community momentum toward open data and tools, these forces enable shared benchmarks—this issue captures and organizes that momentum into emerging standards and best practices.

To accelerate synergistic harmonization of optical, radar (SAR), and lidar satellite observations so land-monitoring products become consistent across sensors, comparable across time and place, and ready for operational use (policy, industry, and practice). The issue focuses on turning heterogeneous measurements into decision-grade, uncertainty-aware maps and time series by featuring studies that explicitly fuse at least two of optical, SAR, and lidar data with rigorous radiometric/structural/temporal harmonization, report uncertainty, and demonstrate cross-region or multi-season generalization.

Toward trustworthy, multimodal sensing for a changing planet, the following is a concise arena of novel methods to synergistically harmonize optical, SAR, and lidar for land monitoring, organized from preprocessing through learning, inference, and uncertainty:

• Focus: algorithms, pipelines, and benchmarks that make cross-sensor, cross-time products consistent, comparable, and operational across regions and seasons.
• Methods: cross-sensor calibration/normalization, co-registration, temporal harmonization, physics-informed and deep-learning fusion, uncertainty estimation, transfer learning, and domain adaptation.
• Data assets: satellite optical (multispectral/hyperspectral), SAR, and spaceborne lidar as primaries; airborne/UAV/in situ data may be used for validation or gap-filling.
• Applications: land cover/land use change; vegetation structure and biomass; agriculture and rangelands; wetlands and peatlands; urban growth and heat; soil moisture/surface deformation; and multi-hazard mapping (e.g., landslides, subsidence).
• Deliverables: algorithms, open datasets/benchmarks, toolchains, reproducible workflows, and comparative evaluations across sites/biomes.
• Cross-sensor radiometric and geometric harmonization: Optical: Pipelines align Landsat and Sentinel-2 band passes, BRDF/view angle, and atmospheric effects to deliver analysis-ready, cross-mission surface-reflectance time series. SAR: Radiometric terrain correction with high-quality DEMs standardizes backscatter across look angles and relief—an essential base layer for fusion.
• Temporal harmonization and gap filling across modalities: SAR-guided cloud removal and sequence models that are temporally consistent with SAR passes.
• Self-supervised shared-latent learning: Contrastive and masked-autoencoding pretraining on paired S1/S2 learning modality-invariant embeddings; multi-objective contrastive schemes further aligning S1/S2 for downstream tasks and cross-modal transfer.
• Physics-aware fusion and joint inversion: Physics-based inversion and radiative/scattering models coupled with data assimilation and ML that can be anchored with sparse GEDI samples.
• Lidar-anchored supervision for wall-to-wall mapping: Use GEDI footprints/metrics as supervision spikes to train models that fuse seasonal optical + SAR into continuous canopy-height/AGB maps, with explicit uncertainty outputs.
• Resolution and modality translation for scale matching: Cross-modal super-resolution and diffusion translators learning optical–SAR mappings to harmonize spatial detail before fusion, while preserving physics via constraints.
• Distribution alignment, domain adaptation and transfer learning: Optimal-transport, adversarial, and teacher–student schemes shrinking modality and domain gaps (sensor, season, region), enabling models trained in one stack/locale to generalize in another.
• Uncertainty-aware harmonization: Conformal prediction (pixel-wise, model-agnostic) and prediction-powered inference wrap fused maps with statistically valid coverage, with explicit calibration, confidence intervals, reliability diagrams, and per-pixel confidence in delivered products.
• Benchmark-driven development and reproducibility: Community multimodal datasets and its cloud-removal/time-series variants, standardize training/validation for cross-sensor fusion, with protocols that enforce benchmarking, comparability, and reproducibility across sites/biomes.

Contributions are warmly invited from researchers and practitioners in the following formats: Original Articles or Letters presenting new fusion algorithms, harmonization pipelines, or operational products with multi-site validation; Reviews or Perspectives synthesizing cross-sensor harmonization, identifying gaps, and proposing roadmaps and standards; and Data/Tools describing curated multi-sensor datasets, benchmarks, or open-source toolchains that lower barriers to adoption. We thank you in advance for your contributions to this Special Issue.

Dr. Ram C. Sharma
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 250 words) can be sent to the Editorial Office for assessment.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-anonymized peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Sensors is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.