Sub-seasonal WRF-Chem reanalysis of extreme Saharan dust outbreaks in spring-summer 2024: balancing phase consistency and aerosol realism

Saharan dust outbreaks intermittently exert strong radiative, air quality and depositional impacts across the Euro-Mediterranean, due to the intrinsic characteristics of this phenomenon, yet their numerical reproduction remains challenging. Here we investigate modelling strategies that preserve spatio-temporal consistency in sub seasonal integrations with WRF-Chem, focusing on three major dust intrusions affecting Italy in 2024: 25 March to 1 April, 18 to 21 June, and 8 to 14 July. We perform a set of reanalysis driven experiments over a single 5 km grid domain spanning North Africa and the Mediterranean into continental Europe, forced by ECMWF IFS analyses at 6 hourly frequencies. Model performance is assessed against complementary observing systems over the Euro-Mediterranean with emphasis on Italy. Our core objective is to quantify how spectral nudging can mitigate large scale phase errors and long run drift, while avoiding an overly constrained mesoscale circulation that may distort dust emission, uplift and transport. In addition, using a sequence of sensitivity runs initialized at increasing lead times, we estimate event dependent spin-up thresholds that stabilize domain integrated dust mass and optical depth, while maintaining realistic emission timing, intensity and extension, to suggest a transferable good practice workflow for episodic dust reanalysis and for longer sub seasonal experiments. Overall, this study frames spectral nudging not as an arbitrary choice but as a tunable constraint whose optimal setting depends on the intended balance between large scale fidelity and internally generated aerosol meteorology feedback, with clear implications for WRF-Chem based dust assessments over Italy and the central western Mediterranean. The focus is on the fact that, despite an approximate 40% increase in computational time, the use of spectral nudging emerges as an optimized approach, both in terms of physical consistency and final computational cost savings. This technique proves particularly advantageous in reducing the overall number of simulations required within the context of sub-seasonal reanalysis.