Monitoring” aims at the assessment selleckchem of the current status of the coastal environment and short term trends, and their (deterministic) short-term forecasts. Such routine analyses and short-term forecasts are required for dealing with all sorts of practical problems such as coastal risk management (coastal flooding and extreme wave conditions), combating ocean pollution (Soomere et al., 2014 and Xi

et al., 2012), search and rescue operations. Similar as with marine spatial planning, monitoring is not a scientific task itself; but, again, the task of monitoring is supported by coastal science in providing methods – in this case, of observations, analysis and prediction. Also, science

is a stakeholder in monitoring efforts as well: Chances to disentangle complex oceanic processes and phenomena are considerably increased if a good state description in space and time is available. For spatial domains and time intervals of practical interest the space–time detailed state of the coastal sea can hardly be determined from observations alone, because a sustainable data acquisition is too expensive. However, amalgamating observations and output of dynamical models enables efficient, consistent and realistic estimations and forecasting of the ocean state (Robinson et al., 1998). selleck kinase inhibitor The challenge of such an amalgamation, also named data assimilation, is the extraction of the most important information from relatively sparse observations, and the propagation of this information in an optimal way into predictive models accounting for errors in the models and observations. There exist still a number of challenges in coastal ocean data assimilation. Diagnostics and metrics for assessing performance of the coastal assimilation models need further improvements.

Coupling between coastal and open-ocean assimilation systems is still an open problem. Obatoclax Mesylate (GX15-070) Forecasting biogeochemistry state in the coastal ocean, although much asked for, is still in infancy. Treatment of river flows, mixing, bottom roughness and small-scale topography is still an issue. Non-homogeneity in space and time of model error statistics needs further consideration. Of particular importance is the optimal use of non-homogeneous data from different origin and platforms. Another application, which is still under development, is the design of observational networks. In numerical “Observation System Simulation Experiments” (OSSEs) possible monitoring networks can be tested, how accurate and efficient field estimates may become, given a certain number or quality of observing stations (Schulz-Stellenfleth and Stanev, 2010). Such OSSEs prepare the ground for designing sustained coastal ocean observing systems, advance the planning and design targeted scientific coastal observations.