High-impact oceanic-atmospheric environmental events, such as El Nino-Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), tropical cyclones (TCs) and Kuroshio large meander (KLM), often induce considerable economic and societal losses on regional or global scales.

Thus, understanding and predicting events of this nature has emerged as a key focus in oceanic and atmospheric research over the last few decades, and is likely to continue to do so into the foreseeable future. For improving the prediction of these events, additional observations with optimal design are urgently needed.

The question is then how to determine such an optimal design, which is precisely the goal of the strategy of targeted observation. Mu Mu (Chinese Academy of Sciences) and coauthors reviewed recent progress in determining the optimal locations for ENSO, IOD, TCs and KLM and discussed their potential role in optimizing observation networks and thus improving forecast skill.

In particular, they suggested an advanced approach of conditional nonlinear optimal perturbation (CNOP) to determine the optimal location for targeted observation. This work, entitled "Target observations for improving initialization of high-impact ocean-atmospheric environmental events forecasting", was published in National Science Review, 2015, Vol 2: 226-236.

This paper briefly reviews recent progress toward targeted observations for improving the prediction of high-impact oceanic-atmospheric environmental events, including ENSO, IOD, TCs and KLM. It is pointed out that current observations are inadequate for providing sufficiently accurate initial conditions for the successful forecasting of these events, and that a targeted observation strategy is urgently needed for the improvement and optimization of observing systems.

Considering the limitations of traditional approaches, new techniques, developed to identify the optimal locations for targeted observations for high-impact events, are introduced. For TCs, the CNOP approach seems superior to other methods in identifying the optimal locations for targeted observations; while for ENSO, the results from different methods are generally similar, verifying the optimal observation locations identified.

Encouraged by the CNOP idea, the optimal initial errors for IOD events were revealed by an ensemble approach, and the optimal locations for targeted observations were identified based on perfect-model predictability experiments using a complex ocean-atmosphere coupled model.

Also introduced are the optimal locations for targeted observations of the KLM, as determined by the CNOP approach. In particular, it is concluded that a targeted observation strategy is an efficient and effective approach to improve the forecast skill of such high-impact events, and will provide guidance for existing and planned observation networks, including TPOS-2020.

It is clarified that, despite the great progress being made, much more work is needed to understand these complex high-impact events and sharpen the tools of targeted observation. Besides, it is pointed out that the success of targeted observation depends upon the performance of numerical models and related assimilation systems.

If a model simulates the high-impact events poorly due to model errors, it cannot be adopted in the implementation of targeted observations for improving the initial conditions and, ultimately, the forecast skill.

Also, if an assimilation system is inadequate, the positive effects of targeted observation cannot be realized. Studies involving targeted observations are therefore challenging. Hence, it is proposed that research on model development, improvements in assimilation systems, and targeted observation should be joined together.

In fact, work in this new field of research is already showing promise, and it is expected that even more exciting advancements are on the horizon. The hope, therefore, is that we will soon see significant improvements in the ability to forecast these high-impact events, which in turn will facilitate progress in disaster prevention, climate change mitigation, and sustainable socio-economic development.