Laser Radar Technology and Applications XXXI

Fathom is an advanced airborne lidar system designed for high-resolution topographic and bathymetric mapping in inland and coastal environments. It integrates a dual-frequency 60 kHz bathymetric laser and a 1.5 MHz topographic channel, achieving an effective sounding rate of 240 kHz in full waveform mode. Validation flights over lakes and rivers using Canadian Hydrographic Service data demonstrate sub-decimeter vertical coherence, reliable water surface detection, and depth penetration up to 32 meters. The system meets IHO S-44 Order 1a and USACE QL0B standards, confirming its suitability for precise, single-pass mapping in diverse aquatic and terrestrial conditions.

NASA’s Mars science community has identified atmospheric wind measurements as a key objective for a future Mars orbiter. To meet this need we previously developed a prototype of the MARLI lidar to measure the height-resolved wind and aerosol profiles. MARLI is a pulsed 1064 nm lidar designed for a polar orbit and to be pointed ~30° from nadir. This allows measuring the height resolved Doppler shift caused by wind in the aerosol backscatter. Although our work on MARLI showed it is practical for space, its mass and power are large given NASA’s new emphasis on smaller Mars missions. We have recently developed a new version called Mini-MARLI. It uses a similar measurement approach but uses a pulsed semiconductor-based laser transmitter that emits near 780 nm and more sensitive detectors. These allow improved wind measurements with lower mass and power. Our measurement model that uses several different atmospheric dust distributions shows that in the lowest 30 km the wind speed errors were 2 to 3 m/s. We are developing a lidar breadboard and plan to use it to demonstrate wind profiling measurements from our laboratory. Details of the lidar design and its measurements will be shown.

Airborne laser scanning with time-of-flight LIDAR has evolved into a key technology for a variety of mapping applications. Determining measurement range requires associating each received echo pulse with its respective emitted laser pulse, but this association is ambiguous if the round-trip time of the pulse exceeds the pulse repetition interval. We introduce a dual-stage approach to this multiple time-around (MTA) ambiguity resolution problem which performs MTA zone assignment not only based on the raw, temporally successive LIDAR measurements, but also in an additional second step after geo-referencing, based on 3D point coordinates. This second stage processes all echoes which are not assigned to a definitive MTA zone by the first stage (standard code-based MTA resolution), and determines the most likely MTA assignment as the one resulting in the highest local point density. This allows introducing information not just from consecutive laser pulses or scan lines, but also from multiple overlapping scan strips. We discuss dual-stage MTA resolution and analyze its advantages over traditional MTA resolution, especially for difficult target situations.

A coordinated wind lidar system was deployed during a multi-month campaign to characterize wind conditions in the Planetary Boundary Layer. Understanding wind conditions in this dynamic layer of the atmosphere is critical to ensure reliable UAS operation. The dual-Doppler lidar system successfully measured resolved wind vectors at low altitudes. A severe weather event occurred during the field campaign and the wind lidars recorded wind dynamics caused by the severe weather.

Point cloud semantic scene understanding is increasingly important in robotics and 3D perception, yet most models are still trained on single-dataset, single-sensor data, limiting their generalization and robustness. We introduce an open-source framework for synergistic training across multiple point cloud datasets and tasks, addressing challenges such as heterogeneous feature inputs and inconsistent label spaces. The framework includes a modular design and separate synergistic task heads for classification, segmentation, and detection, enabling large-scale, multi-task, multi-domain training. By pretraining on diverse datasets spanning varied sensors and environments, our framework fosters the development of 3D models with improved generalization, richer semantic coverage, and robust performance in previously unseen or data-constrained settings.

A single-detector hyperspectral lidar covering the VNIR-SWIR spectral bands was built and used to retrieve the spectral extinction efficiency and the backscattering-to-extinction ratio on a mineral oil aerosol cloud. The system probes 105 wavelengths from 420 nm to 2500 nm by steps of 20 nm. The 1750 nm to 2050 nm water vapour absorption band is probed by steps of 1 nm to estimate the system’s tuning precision and to assess the impact of atmospheric absorption on the system performances.

Mid-Infrared (MIR) gain media have grown in demand in recent years. Since the characteristic peaks in absorption of many organic compounds and atmospheric gases are situated in this region, lasers emitting in this wavelength range have become extremely useful for atmospheric sensing and LADAR applications. Lasing in this wavelength is usually achieved by doping crystals with rare-earth ions that contain the relevant energy transitions. However, due to the low energy associated with these transitions, multi-phonon relaxation (MPR) processes may severely alter the transition rates and make lasing impossible. For this reason, crystal matrices that have low phonon energies are prime candidates for rare-earth doping. In this work, we demonstrate the growth and characterization of crystals that fit these criteria.

This paper presents the development of a COTS 1-mm InGaAs PIN quadrant photodetector for high precision beam tracking in Laser Spot Trackers (LSTs), Precision-Guided Munitions (PGMs), Laser Warning Systems (LWS), and Free Space Optical (FSO) terminal alignment. Targeting defense and aerospace platforms where Size, Weight, Power, and Cost (SWaP-C) are tightly constrained, the module integrates a 1-mm active-area InGaAs PIN photodiode segmented into four quadrants with four application-specific GaAs FET transimpedance amplifiers (TIAs) in a single hermetic TO-can package, delivering a compact solution optimized for high-sensitivity detection for use in mobile or ruggedized defense platforms.

Optimizing the architecture of a NIR optical receiver can significantly enhance detection performance. Achieving this improvement requires a detailed understanding of the photodetector’s intrinsic characteristics—such as gain, excess noise, capacitance - as well as the constraints imposed by the chosen assembly technology, whether implemented with discrete PCB components or through hybrid integration. These factors ultimately define the fundamental noise bandwidth compromise that governs overall receiver sensitivity. This paper examines the impact of key design parameters on InGaAs avalanche photodiode based receivers, outlining how topology, biasing, and front end electronics influence system level performance. Three receiver families are presented, each optimized for specific mission profiles: defend, detect, and communicate. Potential customization paths for enhanced sensitivity, extended bandwidth, or improved environmental robustness are also discussed to guide future development.

In this manuscript a new framework is described for assessing gap fraction and lacunarity of forest canopy from LiDAR derived digital twins. This new approach utilizes ray tracing against the pointcloud derived geometry, allowing for analysis of the effective gap fraction and lacunarity projected along arbitrary lines of sight. This framework is applied to select forest plots to study how effective gap fraction and lacunarity varies with different viewing geometries.

This work introduces DARSoC, a radiation-hardened LiDAR system-on-chip (SoC) integrating laser control, photon detection, analog front-end signal processing, and on-chip histogramming for airborne and spaceborne atmospheric sensing. Designed for applications such as cloud and aerosol profiling and hazard monitoring, DARSoC features a reconfigurable time-of-flight controller with 1–50 m range resolution and a 60 km region-of-interest (ROI) selectable within 700 km. The analog front-end supports both linear-mode and Geiger-mode APDs, combining a low-noise TIA, differential gain, and discriminator with 10-bit TDC-based amplitude and timing extraction. Eight configurable channels enable dynamic range up to 120 dB. Fabricated silicon demonstrates accurate laser timing, selectable ROI, dual-mode detection, and low-noise performance (500 fA/√Hz at 700 MHz). The TDC exhibits <1 LSB noise and ±2 LSB INL, achieving <3 cm range resolution and <1.6 cm precision in altimetry mode. By uniting FPGA-like flexibility with ASIC-level efficiency and radiation tolerance, DARSoC advances compact, high-performance LiDAR for next-generation atmospheric missions.

Frequency-modulated continuous-wave (FMCW) lidar is emerging as a transformative technology in next-generation sensing applications, offering advantages in robustness, precision, and velocity measurement compared to the more commonly used pulsed lidar. Real-world implementations of FMCW lidar vary in terms of modulation methods, spectral usage, detection techniques, receiver architectures, and signal processing depending on the specific application and design requirements. The configurations resulting from combinations of these design elements have varying performance metrics such as range/velocity accuracy, resolution, and sensitivity as well as size, weight, power, cost and complexity differences. This paper presents a review and analysis of FMCW lidar system designs found in literature along with quantitative metrics describing their prevalence.

Foundation models have demonstrated substantial performance gains across domains. In Earth science, Google’s AlphaEarth Foundation (AEF) model has outperformed conventional spaceborne imagery in many applications. Pre-trained on diverse datasets—including Landsat, Sentinel-2, Wikipedia vector layers, GEDI canopy heights, and Copernicus DEM—AEF shows strong capability in 2D classification tasks. However, its utility for 3D applications, particularly integration with airborne laser scanning (ALS) or spaceborne altimetry, remains less explored. This study evaluates AlphaEarth embeddings for predicting terrain elevation and canopy height across three ecoregions. Using an XGBoost regressor, models were trained with ALS or ICESat-2 as reference data. Results indicate that AlphaEarth features alone are insufficient for accurate terrain prediction without additional inputs, despite elevation-related pretraining. In contrast, canopy height predictions show strong standalone performance. Comparisons with multispectral approaches highlight the promise of foundation models, while underscoring the continued need for high-fidelity laser data to calibrate and validate these methods.

Our project is particularly promising amid the proliferation of small, highly maneuverable military drones, for which no effective detection or tracking system exists. A fast hemispherical LiDAR with a five-kilometer radius and an innovative optical spectrometer has been developed. When the laser beam intersects an object, slight scattering occurs, reaching the spectrometer. For small targets or special coatings, the returned signal may reach picowatt levels, while background radiation can be kilowatts—a ratio of about one to 10¹⁵. Under these conditions, the spectrometer extracts the weakest signals, naturally absent and arising only from target-induced scattering. In traditional spectrometers, main parameters are interdependent, preventing ultra-high optimization for specific spectroscopic tasks. Our physical-mathematical model removes this interdependence, enabling ultra-high precision optimization. Advantages include analog operation preventing overload by multiple targets, immunity to radio-electronic jamming, difficulty of location detection by adversaries, and simplicity and compactness. The spectrometer’s mathematical model remains strictly confidential.

We present a landing‑site evaluation system that uses Flash LiDAR point clouds and deep neural networks to identify safe landing zones and hazards for aerial vehicles. Trained exclusively on high‑fidelity synthetic data and validated with NASA’s Morpheus dataset collected via ASC’s Global Shutter Flash LiDAR, our method combines 3D terrain stitching for DEM reconstruction with learned hazard models. All components, including DNN inference and geometric processing, operate in (or near) real time on embedded processors, fulfilling onboard autonomy requirements. We also compare to a classical SLAD algorithm and demonstrate effective simulation‑to‑real transfer for field deployment.

Event-based sensors (EBS) have attracted intense academic, industrial, and defense interest, yet their real-world impact is often limited by event overload in cluttered or highly dynamic scenes. We address this challenge by using EBS as the detector in an active-imaging system, where illumination structure and timing control the event rate at the source. We develop radiometric models and a simple per-pixel event-detection model to compare several feasible event-based active-imaging modalities, highlighting regimes that mitigate overload while preserving EBS advantages (low latency, high dynamic range, low data volume). We then design, build, and test a proof-of-concept system using an amplitude-modulated continuous-wave (AMCW) laser illuminator. Experiments validate the modeling, quantify overload reduction, and map performance–illumination trade-offs, demonstrating that active illumination can make EBS robust and practical for real-world sensing in challenging, high-motion, or high-clutter environments.

Geiger-mode (GmAPD) Lidar sensors are comprised of 2D arrays of avalanche photodiodes that are record single photon arrivals with 1-2ns jitter. These GmAPD sensors can operate in otherwise untenable SNR regimes, utilizing coincidence processing algorithms to recover the 3D terrain from the noisy measurements. However, these sensors are also susceptible to crosstalk, where the pixels avalanches trigger neighboring pixels to also avalanche due to internally scattering photons. Unlike dark current, these phantom points are correlated with the signal and do not average out, blurring the derived imagery downrange and perpendicular to the line of sight. In this work, we utilize synthetic GmAPD LIDAR point clouds to characterize the effect of crosstalk points on 3D imagery. We evaluate these point clouds both qualitatively and quantitatively, comparing the changes in standard resolution measurements due to the increasing presence of crosstalk points. We also investigate what terrain features are most corrupted by crosstalk noise in these point clouds. We discuss how these various findings could allow crosstalk noise to be addressed from a data-based perspective.

In this manuscript, an end-to-end processing chain for Geiger-mode LiDAR at the edge is described. This chain includes GPU accelerated algorithms for pointcloud synthesis, coincidence processing, registration and secondary light detection to take advantage of the additional computational power afforded GPU systems on a chip. This processing chain is applied to data collected with the PHOENIX High CASTLE Geiger-mode LiDAR sensor, and an analysis of how performance scales with system resource allocation is presented, including a discussion of computational, data fidelity, and SWaP tradeoffs.