Assimilating bio-optical glider data during a phytoplankton bloom in the southern Ross Sea

Abstract

Abstract. The Ross Sea is a region characterized by high primary productivity in comparison to other Antarctic coastal regions, and its productivity is marked by considerable variability both spatially (1-50 km) and temporally (days to weeks). This variability presents a challenge for inferring phytoplankton dynamics from observations that are limited in time or space, which is often the case due to logistical limitations of sampling. To better understand the spatiotemporal variability in Ross Sea phytoplankton dynamics and to determine how restricted sampling may skew dynamical interpretations, high-resolution bio-optical glider measurements were assimilated into a one-dimensional biogeochemical model adapted for the Ross Sea. The assimilation of data from the entire glider track using the micro-genetic and local search algorithms in the Marine Model Optimization Testbed improves the model-data fit by ∼ 50 %, generating rates of integrated primary production of 104 g C m −2 yr −1 and export at 200 m of 27 g C m −2 yr −1 . Assimilating glider data from three different latitudinal bands and three different longitudinal bands results in minimal changes to the simulations, improves the model-data fit with respect to unassimilated data by ∼ 35 %, and confirms that analyzing these glider observations as a time series via a one-dimensional model is reasonable on these scales. Whereas assimilating the full glider data set produces well-constrained simulations, assimilating subsampled glider data at a frequency consistent with cruise-based sampling results in a wide range of primary production and export estimates. These estimates depend strongly on the timing of the assimilated observations, due to the presence of high mesoscale variability in this region. Assimilating surface glider data subsampled at a frequency consistent with available satellite-derived data results in 40 % lower carbon export, primarily resulting from optimized rates generating more slowly sinking diatoms. This analysis highlights the need for the strategic consideration of the impacts of data frequency, duration, and coverage when combining observations with biogeochemical modeling in regions with strong mesoscale variability.

Enhancing Ocean Biogeochemical Models With Phytoplankton Variable Composition

“…The cost function is computed as the weighted average of the squared difference between observed and simulated values, using the sample standard deviation of all observations of type m,σ m , to appropriately weight the contribution of each model-data misfit, as in previous studies (e.g., Friedrichs et al, 2006;Hemmings and Challenor, 2012;Kaufman et al, 2018). However, because the variables that we compared tend to be distributed log normally, especially at depths where their concentration are high, we use a square-root transformation to calculate the model data difference (Dadou et al, 2004;Hemmings and Challenor, 2012).…”

Drivers of the spatial phytoplankton gradient in estuarine–coastal systems: generic implications of a case study in a Dutch tidal bay

“…The standing stock of phytoplankton is a function of sources and sinks that are subject to both biotic and abiotic influences (Lancelot and Muylaert, 2011;Jiang et al, 2015). Phytoplankton growth is regulated by bottom-up factors such as nutrients, light, and temperature (Underwood L. Jiang et al: Drivers of the spatial phytoplankton gradient in estuarine-coastal systems and Kromkamp, 1999;Cloern et al, 2014), while natural mortality and grazing pressure from zooplankton, suspension feeders, and other herbivores contribute to the loss of phytoplankton biomass (Kimmerer and Thompson, 2014). Physical transport can act as either a direct source or sink, driving algal cells into or out of a certain region (Martin et al, 2007;Qin and Shen, 2017).…”

“…Nitrate can support more phytoplankton biomass in microtidal estuaries than in macrotidal estuaries (Monbet, 1992). The relative importance of zooplankton and bivalve grazing on phytoplankton varies spatially Herman et al, 1999;Kimmerer and Thompson, 2014). These complexities make it challenging to discern the driving mechanisms of the spatial phytoplankton gradient, and comparative studies of different systems are lacking (Kromkamp and van Engeland, 2010;Cloern et al, 2017).…”

“…In situ observation, remote sensing, and numerical modeling are common techniques to reveal spatial patterns and detect their biophysical controls (Banas et al, 2007;Grangeré et al, 2010;Srichandan et al, 2015). Shipboard measurements of chlorophyll a (chl a) provide a precise and dynamic assessment of the phytoplankton variability; however, the temporal (usually monthly) and spatial (usually tens of kilometers) resolutions are limited compared with satellite images and numerical models (Soetaert et al, 2006;Valdes-Weaver et al, 2006;van der Molen and Perissinotto, 2011;Cloern and Jassby, 2012;Kaufman et al, 2018). Remote sensing of chl a reveals the surface distribution with a suf-ficiently high spatial resolution and coverage, although only under favorable weather conditions (Srichandan et al, 2015).…”

How Critical Is the Assimilation Frequency of Water Content Measurements for Obtaining Soil Hydraulic Parameters with Data Assimilation?

“…p. 2 of 10 frequency, duration, and coverage on DA results was noted, for instance, when observations were combined with biogeochemical modeling (Kaufman et al, 2018), when the data were assimilated for real-time flood forecasts (Mazzoleni et al, 2017), and when the land DA system was used to couple the microwave remote sensing and land surface model and then improve the accuracy of land surface fluxes and status estimation (Lu et al, 2016). Results of using geophysical and remote sensing DA to estimate soil water contents depended on the assimilation frequency (Cosenza, 2016;Rosolem et al, 2014).…”