The assimilation of satellite‐derived data into a one‐dimensional lower trophic level marine ecosystem model

Abstract

Lower trophic level marine ecosystem models are highly dependent on the parameter values given to key rate processes, however many of these are either unknown or difficult to measure. One solution to this problem is to apply data assimilation techniques that optimize key parameter values, however in many cases in situ ecosystem data are unavailable on the temporal and spatial scales of interest. Although multiple types of satellite-derived data are now available with high temporal and spatial resolution, the relative advantages of assimilating different satellite data types are not well known. Here these issues are examined by implementing a lower trophic level model in a one-dimensional data assimilative (variational adjoint) model testbed. A combination of experiments assimilating synthetic and actual satellite-derived data, including total chlorophyll, size-fractionated chlorophyll and particulate organic carbon (POC), reveal that this is an effective tool for improving simulated surface and subsurface distributions both for assimilated and unassimilated variables. Model-data misfits were lowest when parameters were optimized individually at specific sites; however, this resulted in unrealistic overtuned parameter values that deteriorated model skill at times and depths when data were not available for assimilation, highlighting the importance of assimilating data from multiple sites simultaneously. Finally, when chlorophyll data were assimilated without POC, POC simulations still improved, however the reverse was not true. For this twophytoplankton size class model, optimal results were obtained when satellite-derived size-differentiated chlorophyll and POC were both assimilated simultaneously.

Using biogeochemical data assimilation to assess the relative skill of multiple ecosystem models in the Mid-Atlantic Bight: effects of increasing the complexity of the planktonic food web

“…Each of the five marine ecosystem models were embedded in a 1-D (vertical) physical model that contains standard parameterizations for vertical advection, diffusion and sinking particles that have been thoroughly described in a number of other 1-D modeling studies Ward et al, 2010;Xiao and Friedrichs, 2014 the same as in Xiao and Friedrichs (2014), i.e., provided by a corresponding three-dimensional (3-D) 1P1Z model implementation (Hofmann et al, 2008;. Models with two size classes were initialized as one-half of the 3-D 1P1Z concentrations, and models with three size classes were initialized as one-third of these concentrations.…”

“…Based on the results of Xiao and Friedrichs (2014), three types of daily 9km data were derived from the Sea-viewing Wide Field-of-view Sensor (SeaWiFS) and assimilated into the five models described above (Table 2): size-fractionated chlorophyll a (Pan et al, 2010), total chlorophyll a (computed as the sum of the size-fractionated chlorophyll) and particulate organic carbon (Stramska and Stramski, 2005). Although these satellite data were all derived using empirical or semi-analytical algorithms, they have demonstrated considerable success in their agreement with in situ data.…”

“…The specifics of the optimization implementation are well documented in Xiao and Friedrichs (2014), and thus only a brief description of the key properties of the variational adjoint data assimilative framework is provided here. The variational adjoint method is a nonlinear, weighted least-squares optimization method that minimizes the misfit between the model estimates and the observational data by optimizing a subset of model parameters (e.g., Lawson et al, 1995Lawson et al, , 1996.…”

“…The choice of parameters for optimization depends strongly on the data available for optimization. When size-differentiated chlorophyll and particulate organic carbon data are available for assimilation, Xiao and Friedrichs (2014) determined that successful assimilation results are obtained as long as data from multiple sites are assimilated, and the subset of parameters to be optimized includes maximum chlorophyll : carbon (Chl : C) ratios, maximum phytoplankton growth rates and zooplankton basal metabolism rates. Because each optimized parameter is size specific (that is, each phytoplankton size class has a distinct Chl : C ratio and growth rate) and each zooplankton size class has a distinct basal metabolism rate (Table 1), the number of optimized parameters increases with increasing model complexity.…”

Biological data assimilation for parameter estimation of a phytoplankton functional type model for the western North Pacific

“…Kuroda and Kishi, 2004;Fiechter et al, 2013;Toyoda et al, 2013;Xiao and Friedrichs, 2014). Shigemitsu et al (2012) applied a unique assimilative approach to an LTL marine ecosystem model, using a micro-genetic algorithm (µ-GA) (Krishnakumar, 1990).…”