Tiedje, B, Moll, A, Kaleschke, L (2010). Comparison of temporal and spatial structures of chlorophyll derived from MODIS satellite data and ECOHAM3 model data in the North Sea. JOURNAL OF SEA RESEARCH, 64(3), 250-259.
In this paper, we assess the ability of the three-dimensional physical-biological coupled model ECOHAM3 to reproduce the seasonal dynamics and the spatial distribution of chlorophyll in the greater North Sea area. We use chlorophyll products from MODIS and in situ measurements of chlorophyll from long-term monitoring as ground truths to validate the spatial and temporal phytoplankton dynamics for the year 2004. The MODIS diffusive attenuation coefficient showed that the depth up to which the chlorophyll concentration is integrated for the MODIS chlorophyll product seldom exceeds 10 m. The upper 10 m of the water column are represented by the top layer of the model. Monthly regional distributions show a large-scale overestimation by the model while 87% of the difference values lie in the interval of - 1 mg/m(3) and 1 mg/m(3). Only in the coastal regions the model underestimates the satellite data. For comparison of annual cycles and spatial gradients we used in situ observations to check the reliability of satellite and model data. On a transect through the North Sea the best agreement between the three data sets is achieved in the central and northern North Sea during winter and autumn. All data sets show an increase at the northern part of this transect during spring and summer. Averaged over the whole year the satellite data agrees well with in situ data. The model data exceeds the in situ and satellite data along the transect by 50% except for the coastal region where the in situ and satellite data have a much steeper slope and finally exceed the model data. The monthly mean values at a stratified station in the northern North Sea for both model and satellite data show a spring and autumn bloom and lie within the standard deviation of the in situ data during spring time. In the well mixed German Bight all three data sets show an increase of the chlorophyll concentration until April. In summer the data sets have different developments within the observed variability. An analysis of a Gaussian four-parameter fit to the weekly satellite and model time series describes the phytoplankton dynamics in stratified waters quantitatively. By considering the 576 (24 x 24 pixels) different annual chlorophyll cycles of the satellite data in a region of 1 degrees by 1 degrees, the extremely different biological evolution during spring bloom on such a high resolution becomes apparent, which is not reproduced by the model. The best agreement between the model and satellite data is achieved in the timing of the chlorophyll maximum concentration. This comparison of data sets is the first validation of the model with satellite data. (C), 2010 Elsevier B.V. All rights reserved.