*Article* **SARS-CoV-2 Spread Forecast Dynamic Model Validation through Digital Twin Approach, Catalonia Case Study**

**Pau Fonseca i Casas 1,\*, Joan Garcia i Subirana 1, Víctor García i Carrasco <sup>1</sup> and Xavier Pi i Palomés <sup>2</sup>**


**Abstract:** The spread of the SARS-CoV-2 modeling is a challenging problem because of its complex nature and lack of information regarding certain aspects. In this paper, we explore a Digital Twin approach to model the pandemic situation in Catalonia. The Digital Twin is composed of three different dynamic models used to perform the validations by a Model Comparison approach. We detail how we use this approach to obtain knowledge regarding the effects of the nonpharmaceutical interventions and the problems we faced during the modeling process. We use Specification and Description Language (SDL) to represent the compartmental forecasting model for the SARS-CoV-2. Its graphical notation simplifies the different specialists' understanding of the model hypotheses, which must be validated continuously following a Solution Validation approach. This model allows the successful forecasting of different scenarios for Catalonia. We present some formalization details, discuss the validation process and present some results obtained from the validation model discussion, which becomes a digital twin of the pandemic in Catalonia.

**Keywords:** SARS-CoV-2; COVID-19; SEIRD (Susceptible, Exposed, Infected and Recovered and Death); SDL; Catalonia
