Reproducibility and Replicability in SAR Remote Sensing

Abstract

Modern science is built on systematic experimentation and observation. Today's form of communicating scientific results with written articles is the foundation we build our work on. The reproducibility and replicability of the experiments and observations are central to this process of validation. However, reproducibility and replicability are not always guaranteed, sometimes referred to as “crisis of reproducibility.” We believe that remote sensing, in general, suffers from this crisis. To analyze the extent of the crisis, we conducted a survey on the state of reproducibility in remote sensing. Based on this survey, we map the problem of reproducibility with a focus on synthetic aperture radar remote sensing, as this is our area of research. We also give advice on how to improve reproducibility in remote sensing.

Keywords

• Remote sensing
• replicability
• reproducibility
• survey