She et al. 
Integrated Coastal and Biological Observing 
Frontiers In Marine Science | www.frontlersln.org 
3 
July 2019 I Volume 6 | Article 314 
FIGURE 1 | Integrated observing - unlocking the value of ocean observing by Integrating observations In three dimensions: flt-for-purpose, parameter, and 
Instrumental. 
ship observations from the offline monitoring programs, the 
data gaps for operational forecasting and interim reanalysis can 
be largely filled. However, the difficulty of harmonizing multi 
networks should not be underestimated, in which significant 
institutional and community barriers should be overcome. The 
cost-effectiveness of the observing can be improved by optimal 
sampling strategy design, including cost-benefit analysis of 
the monitoring technology. Many sampling strategy design 
studies have been carried out, using methods ranging from 
statistical design e.g., Springtall and Meyers (1991), She et al. 
(2006), and Alvarez and Mourre (2012) to Observing System 
Simulation Experiments (OSSEs, Oke et al., 2015; She et al., 
2017). However, these optimal sampling design studies were 
mainly dedicated for operational forecasting and reanalysis. 
Few of them have included cost-benefit analysis and fit-for- 
multi-purpose optimization. It should be noted that a significant 
amount of new knowledge and new observations will be needed 
for the optimization, which constitute the third stage of the 
implementation. 
Parameter Integration 
Fit-for-purpose integration improves observation adequacy, 
appropriateness, and cost-effectives. However, the required 
observations also have to be easily accessible by the users. In 
many cases, data exist but not available as they are managed by 
different sectorial data centers and also subjected to different data 
policies. This makes data sharing more difficult and data usage 
less efficient. Integration of marine observations across entire 
parameter spectrum can significantly improve the efficiency 
of the data use. 
In Europe, the EMODnet (Miguez et al., this issue) is 
dedicated to integrate marine data across a full parameter 
spectrum - bathymetry, biology, chemistry, coastal mapping, 
geology, human activity, and physics. Recently emerging 
variables e.g., riverine inputs, underwater noise, sediment grain 
size, marine litter, and other datasets have been added in the 
portals. It was found, by the EMODnet Sea Basin Checkpoint 
projects, that the high integration level of marine data, such as 
done by EMODnet, has greatly facilitated the user applications 
and unlocks the value of observations. 
Instrumental Integration 
The value of observations can only be realized when they 
are used. In situ observing (including sensor technology and 
sampling schemes), remote sensing, and modeling are three 
ways of tracking ocean conditions. The instrumental integration 
means to produce needed products by integrating these three 
tools, e.g., data assimilation. Although such integration has been 
developed for decades, most of the operational assimilation just 
started in this century and mainly for open oceans and for 
physical variables. In Europe, the most well-known instrumental 
integration effort is Copernicus Marine Environment Monitoring 
Service (CMEMS, Le Traon et al., this issue). The lack of 
integration in coastal ocean and biogeochemical variables may 
be attributed to several reasons, e.g., lack of efficient schemes 
assimilating high frequency and multi-scale coastal observations, 
lack of skills in models to resolve fine scale features and 
biogeochemical processes and lack of qualified observations. 
These issues are major challenges in the instrumental integration 
of the coastal observing system, which should be resolved in 
the future. 
BIOLOGICAL OBSERVATIONS 
Biological ocean observations are any data collected in a 
systematic and regular basis which are based on living 
ocean inhabitants.