Browsed by
Category: forecast visualization

ZOOMing into the Age of Digital Collaborations

ZOOMing into the Age of Digital Collaborations

Contributed by Bart van Osnabrugge, Antara Dasgupta, Louise Arnal, Rebecca Emerton and Shaun Harrigan Are physical meetings strictly necessary to foster effective collaborations?  The Context Hearing the laments about online conferences, limited interactions and zoom fatigue, it seems easy to go with the answer “yes they are”. Yes, meeting in person is fun and makes connecting a lot easier. There is no rivalling going out for dinner or partying after a full day of presentations. However, the virtual world also offers…

Read More Read More

How can we verify the predictive performance of ensemble hydroclimatic forecasts?

How can we verify the predictive performance of ensemble hydroclimatic forecasts?

Contributed by Zeqing Huang and Tongtiegang Zhao This blog aims to contribute to the large scientific discussion on the performance assessment of ensemble hydroclimatic forecasts. We are particularly driven by the valuable global precipitation and temperature forecasts generated by global climate models (GCMs) (Pappenberger and Buizza, 2009; Kirtman et al., 2014; Bauer et al., 2015; Becker et al., 2020; Crochemore et al., 2021). Their forecasts have been widely used in hydrological modeling and water resources management, including flood warning  (Alfieri…

Read More Read More

Representative Direct Ensemble Uncertainty Visualizations: Conveying Uncertainty Using A Small Portion of The Data

Representative Direct Ensemble Uncertainty Visualizations: Conveying Uncertainty Using A Small Portion of The Data

Contributed by Le Liu, PhD., School of Computer Science Northwestern Polytechnical University.   As we know, ensemble approaches are widely adopted to estimate forecasts uncertainty. In atmospheric sciences, these approaches are specifically categorized into two types: multi-model ensembles and perturbed parameter ensembles. The former runs multiple numerical prediction models with the same initial parameters to estimate the atmospheric evolution, while the latter runs a single model multiple times with slightly perturbed initial conditions. They are usually combined to form the final forecast….

Read More Read More