Singular vectors in atmospheric sciences: A review

During the last decade, singular vectors (SVs) have received a lot of attention in the research and operational communities especially due to their use in ensemble forecasting and targeting of observations. SVs represent the orthogonal set of perturbations that, according to linear theory, will grow fastest over a finite‐time interval with respect to a specific metric. Hence, the study of SVs gives information about the dynamics and structure of rapidly growing and finite-time instabilities representing an important step toward a better understanding of perturbations evolution in the atmosphere. This paper reviews the SV formulation and gives a brief overview of their recent applications in atmospheric sciences. A particular attention is accorded to the SV sensitivity to different parameters such as optimization time interval, norm, horizontal resolution and tangent linear model, various choices leading to different initial structures and evolutions.