Operational flood forecasting using spatially distributed hydrological models requires accurate, high-resolution gridded meteorological forcing delivered in near real time, particularly for systems operating on hourly update cycles. To support this application, this study evaluates four spatial interpolation methods—inverse distance weighting (IDW), ordinary kriging (OK), external drift kriging (EDK), and a Gaussian copula-based approach—for producing high-resolution gridded meteorological fields across Germany. Performance is assessed using station-based cross-validation and benchmarking against high-resolution reference datasets from the German Weather Service (DWD). Results show that hourly precipitation exhibits clear climatological variogram structures but limited instantaneous spatial coherence. Consequently, IDW and OK show very similar performance for most situations, with OK providing only modest improvements during high-intensity or more spatially coherent events. For daily totals derived from the hourly fields, both methods substantially outperform RADOLAN and show close agreement with HYRAS. In contrast to precipitation, temperature fields exhibit stronger spatial coherence and smoother spatial structure, making them particularly well suited to multivariate dependence modelling. Consequently, the Gaussian copula approach captures spatial patterns more accurately than IDW and kriging-based methods, resulting in substantial improvements in interpolation accuracy. Evaluation against the hourly reference grids of the German Weather Service (DWD) shows that the copula method reliably reproduces large-scale spatial patterns and achieves an RMSE of 1.19 °C over 1995–2024. These results are obtained without explicitly accounting for land-use-dependent variables, highlighting the efficiency of the approach for near-real-time applications. To examine suitability for operational use, the methods are further tested under extreme conditions, with case studies centred on the July 2021 flood and the summer 2024 flood events. Two recent heavy-rainfall events associated with flooding (Ahr 2021 and Fils 2024) further show that OK captures the spatial structure of extreme precipitation slightly better than the other methods, although IDW performs similarly and both substantially outperform RADOLAN. Overall, the study shows that hourly precipitation interpolation remains fundamentally constrained by weak event-scale spatial dependence and station density, which limits the benefits of more advanced methods, whereas temperature fields benefit strongly from copula-based modelling. Method choice should therefore depend on the variable of interest: for hourly precipitation, advanced interpolation provides only limited gains over simpler approaches, whereas for temperature, copula-based modelling offers clear advantages.