Chinchella et al. (2026) On the accuracy of optical disdrometer measurements

Identification

• Journal: Atmospheric Research
• Year: 2026
• Date: 2026-02-14
• Authors: Enrico Chinchella, Arianna Cauteruccio, Luca G. Lanza
• DOI: 10.1016/j.atmosres.2026.108865

Research Groups

• Department of Civil, Chemical and Environmental Engineering, University of Genova, Genoa 16145, Italy
• WMO Measurement Lead Centre “B. Castelli” on Precipitation Intensity, Genoa 16145, Italy

Short Summary

This study systematically quantifies the instrumental bias of two widely used optical disdrometers (OTT Parsivel2 and Thies LPM) in a laboratory setting using a traceable raindrop generator, revealing significant underestimation of drop size and integral properties, and biases in fall velocity, highlighting the critical need for rigorous calibration.

Objective

• To traceably and systematically quantify the instrumental bias of optical disdrometer measurements (OTT Parsivel2 and Thies LPM) in a laboratory setting.
• To assess the measurement accuracy and potential biases in individual drop size and fall velocity, and their propagation to derived integral properties of precipitation (rainfall intensity, kinetic energy, radar reflectivity).

Study Configuration

• Spatial Scale: Laboratory experiments conducted at the Precipitation Measurement Laboratory of the University of Genova. The study involved two optical disdrometers: OTT Parsivel2 (30 mm laser beam width) and Thies LPM (20 mm laser beam width). Water drops with equivolumetric diameters ranging from 0.6 mm to 5 mm were generated and released from a height of 1.2 m above the instrument's sensing area. Metallic spheres with diameters from 1 mm to 5 mm were released from 0.37 m.
• Temporal Scale: Individual drop measurements were captured by a Photogrammetric Verification System (PVS) with flashes at 2.5 millisecond intervals. Disdrometer data were analyzed from raw measurements and, where available, from 1-minute aggregated Particle Size and Velocity Distribution (PSVD) matrices.

Methodology and Data

• Models used:
  • Analytical model based on Rayleigh's (1879) and Lamb's (1932) work, as adapted by Zhang et al. (2019), to simulate and fit drop oscillation behavior.
  • Marshall–Palmer equation (Marshall and Palmer, 1948) for Drop Size Distribution (DSD) in Monte Carlo simulations to assess integral properties.
  • Equations for calculating integral precipitation properties: Rainfall Intensity (RI), Kinetic Energy flux (E), and Radar Reflectivity (Z) from measured PSVD.
• Data sources:
  • Calibrated Rainfall Generator (CRG) developed at the University of Genova, capable of producing single water drops with diameters from 0.6 mm to 5.5 mm.
  • Photogrammetric Verification System (PVS) consisting of a 24-megapixel APS-C CMOS camera with a macro lens, providing a maximum spatial resolution of 3.89 µm, used for traceable reference measurements of drop diameter and fall velocity.
  • Calibrated steel spheres with nominal diameters ranging from 1 mm to 5 mm.
  • Raw measurement data and PSVD matrix outputs from two commercial optical disdrometers: OTT Parsivel2 and Thies Laser Precipitation Monitor (LPM).

Main Results

• Calibration using metallic spheres yielded significantly different results compared to water drops, demonstrating that artificial particles are not appropriate for calibrating optical disdrometers.
• Drop Size Measurement Bias:
  • Both instruments significantly underestimate drop size by up to 20% on average.
  • The Thies LPM, when using its internal post-processing algorithm (PSVD matrix), showed peaks of underestimation up to 40%.
  • The OTT Parsivel2 exhibited a bias ranging from +5% (overestimation for 4 mm and 5 mm drops) to -10% (underestimation for smaller drops).
  • The mean bias and dispersion of drop size measurements varied along and across the laser beam, with differences up to 30% observed for the Thies LPM between opposite edges of the beam.
• Fall Velocity Measurement Bias:
  • The OTT Parsivel2 generally overestimated fall velocity by approximately 5% to 10%.
  • The Thies LPM generally underestimated fall velocity by up to 15%.
  • Dispersion in fall velocity measurements increased with increasing drop diameter for both instruments.
  • For the Thies LPM, fall velocity was significantly overestimated near the transmitter head and underestimated near the receiver head along the laser beam.
• Integral Properties Measurement Bias:
  • Integral properties (Rainfall Intensity, Kinetic Energy flux, Radar Reflectivity) were largely underestimated, particularly at low rainfall intensity, with performance improving as intensity increased.
  • The OTT Parsivel2 underestimated Rainfall Intensity by about 10% to 15%.
  • The Thies LPM showed a much larger underestimation of Rainfall Intensity, approximately three times that of the OTT Parsivel2.
  • The OTT Parsivel2 underestimated Kinetic Energy flux by about 5% at lower intensities, approaching zero bias at higher intensities.
  • The Thies LPM strongly underestimated Kinetic Energy flux by up to 50% at the lowest rainfall intensities.
• Drop oscillation (eccentricity) was found to have no significant impact on the measured drop size or fall velocity bias, suggesting the instruments can account for variations in drop shape.
• False readings were observed: "ghost drops" (up to 1 mm diameter) for the OTT Parsivel2 when large drops fell near the receiver head, and interference from speedlight flashes for the Thies LPM.

Contributions

• This study provides the first traceable and systematic quantification of instrumental bias for widely used optical disdrometers in a controlled laboratory environment.
• It demonstrates the inadequacy of using artificial particles (metallic spheres) for calibrating optical disdrometers, highlighting the necessity of using actual water drops.
• The research quantifies the propagation of individual drop measurement biases to derived integral precipitation properties, offering crucial insights for applications like radar calibration and hydrological modeling.
• It underscores the importance of laboratory calibration for non-catching precipitation instruments and supports the need for standardized calibration procedures, such as the emerging European Standard (CEN, 2025).
• The findings reveal significant undisclosed filtering and/or post-processing in commercial disdrometers, advocating for greater transparency from manufacturers.

Funding

• EURAMET project 18NRM03, ‘INCIPIT: Calibration and Accuracy of Non-Catching Instruments to Measure Liquid/Solid Atmospheric Precipitation’, funded by the EMPIR programme (co-financed by Participating States and the European Union's Horizon 2020 research and innovation programme).
• Italian National PRIN2022MYTKP4 project, ‘Fostering innovation in precipitation measurements: from drop size to hydrological and climatic scales’.

Citation

@article{Chinchella2026accuracy,
  author = {Chinchella, Enrico and Cauteruccio, Arianna and Lanza, Luca G.},
  title = {On the accuracy of optical disdrometer measurements},
  journal = {Atmospheric Research},
  year = {2026},
  doi = {10.1016/j.atmosres.2026.108865},
  url = {https://doi.org/10.1016/j.atmosres.2026.108865}
}