Rain Field Retrieval by Ground-Level Sensors of Various Types

H Messer, A Eshel, HV Habi, Shay Sagiv, X Zheng

Research output: Contribution to journalReview articlepeer-review


Rain gauges (RGs) have been utilized as sensors for local rain monitoring dating back to ancient Greece. The use of a network of RGs for 2D rain mapping is based on spatial interpolation that, while presenting good results in limited experimental areas, has limited scalability because of the unrealistic need to install and maintain a large quantity of sensors. Alternatively, commercial microwave links (CMLs), widely spread around the globe, have proven effective as near-ground opportunistic rain sensors. In this study, we study 2D rain field mapping using CMLs and/or RGs from a practical and a theoretical point of view, aiming to understand their inherent performance differences. We study sensor networks of either CMLs or RGs, and also a mixed network of CMLs and RGs. We show that with proper preprocessing, the rain field retrieval performance of the CML network is better than that of RGs. However, depending on the characteristics of the rain field, this performance gain can be negligible, especially when the rain field is smooth (relative to the topology of the sensor network). In other words, for a given network, the advantage of rain retrieval using a network of CMLs is more significant when the rain field is spotty.
Original languageEnglish
Article number877336
Number of pages11
JournalFrontiers in Signal Processing
StatePublished - 17 Aug 2022


  • Cramer-Rao bound
  • Commercial microwave links
  • Rain estimation
  • Sensor network
  • Spatial interpolation


Dive into the research topics of 'Rain Field Retrieval by Ground-Level Sensors of Various Types'. Together they form a unique fingerprint.

Cite this