Skill assessment of numerical models using satellite remote sensing data

Complex numerical models of the Earth's environment, based around 3-D or 4-D time and space domains are routinely used for a large number of applications. In the context of marine ecosystems, models are used for a number of applications including climate predictions, fishery and coastal management and environmental impact assessment. As these models increase in complexity and our reliance on them increases, so does the need to assess the accuracy of their predictions.

Satellite remote sensing is a unique tool to assess model skill at various spatial and temporal scales. NEODAAS provides a range of products that can be used to evaluate model outputs such as sea surface temperature, chlorophyll-a concentration or primary production.

Operational evaluation of the UK Meteorological Office Forecasting Ocean Assimilation Model over the Atlantic-European North West shelf is performed by NEODAAS within the context of the National Centre for Ocean Forecasting (http://www.neodaas.ac.uk/multiview/uk/metoffice_model) using recommendations and techniques from Shutler et al. (2011) and Saux Picart et al. (2012).

Figure 1 shows an example of model validation, using Saux Picart et al. (2012) wavelet analysis technique.

Figure 1: (a) Wavelet analysis performed on monthly chlorophyll concentration field from Polcom/ERSEM model and (b) derived from MODIS sensor for April 2004 (c) represents the skill score on a range/scale plot.

Saux Picart, S., Butenschon, M., & Shutler, J. (2012) Wavelet-based spatial comparison technique for analysing and evaluating two-dimensional geophysical model fields Geoscientific Model Development, 5, 1-8, doi:10.5194/gmd-5-1-2012.

Shutler, J., Smyth, T., Saux Picart, S., Wakelin, S., Hyder, P., Grant, M., Orekhov, P., Tilstone, G., & Allen, J.I. (2011) Evaluating the ability of a hydrodynamic ecosystem model to capture inter- and intra-annual spatial characteristics of chlorophyll-a in the north east Atlantic Journal of Marine Systems, 88, 169-182, 10.1016/j.jmarsys.2011.03.013.