Abstract

We introduce a parsimonious and flexible subclass of the Closed Skew-Normal (CSN) distribution and derive several useful properties that make it particularly well-suited for modeling spatial random fields. Within this framework, we investigate the behavior of least squares estimators. Furthermore, leveraging the Fisher information matrix under a specific parametrization, we explore the use of the Kullback–Leibler (KL) divergence to compare this subclass with the Gaussian case, thereby highlighting the influence of skewness. A comparative simulation study is also presented, incorporating different levels of bias to assess the robustness and performance of the proposed model.