Assessing the economic viability of Argo float recovery: A probabilistic, float-type specific analysis under operational uncertainty

Abstract. The Argo program provides sustained global ocean observations through a large array of autonomous profiling floats. As the system evolves towards deep and biogeochemical observations, the economic implications of maintaining and renewing the array have become increasingly important. While Argo profiling floats operating in the open ocean are traditionally not recovered at the end of their lifetime, recent initiatives motivated by the challenges of operating floats in the shallower European marginal seas have demonstrated the technical feasibility of recovery, refurbishment, and relaunch. However, the economic value of such strategies remains poorly quantified under realistic operational uncertainty. This study evaluates the economic value of the recovery of Argo floats using a probabilistic framework that explicitly accounts for uncertainty in recovery costs and recovery success rates. Recovery and non-recovery strategies are compared for five representative float categories: Core, Deep, and three BGC configurations differing in sensor payload complexity. A stepwise methodology is applied, combining deterministic baseline analysis, sensitivity analysis, mission-level cost-effectiveness analysis, and Monte Carlo simulations. Results reveal clear differences in the economic performance of float recovery across float types. Recovery reduced the estimated cost per additional profile by approximately 57 % for Core floats and 77 % for BGC floats, indicating that redeployment can substantially increase observing efficiency. For Core floats, recovery is economically favorable in most simulated scenarios, although outcomes are more sensitive to recovery costs and success rates than for other float types. In contrast, Deep floats and all three BGC configurations analyzed in this study (BGC (O₂), BGC (O₂, FLBB), and BGC (+3)) exhibit very high probabilities of economic viability under the assumptions adopted in the Monte Carlo analysis, with the greatest economic margins observed for the most instrumented BGC platforms. Core floats remain economically viable in approximately 95 % of simulated scenarios. Overall, our findings demonstrate that the economic robustness differs substantially among float types, and that probabilistic analysis is essential for supporting efficient and evidence-based decision-making for more sustainable ocean observing systems. By explicitly accounting for cost uncertainty, the proposed framework provides quantitative guidance for designing recovery strategies that enhance the efficiency and resilience of the evolving Argo array.