Carla Caverhill

Regionally and seasonally differentiated primary production in the North Atlantic

A bio-geochemical classification of the N. Atlantic Basin is presented according to which the basin is first divided into four primary algal domains: Polar, West-Wind, Trades and Coastal. These are in turn sub-divided into smaller provinces. The classification is based on differences in the physical environment which are likely to influence regional algal dynamics. The seasonally-differentiated parameters of the photosynthesis-light curve (P-I curve) and parameters that define the vertical structure in chlorophyll profile are then established for each province, based on an analysis of an archive of over 6000 chlorophyll profiles, and over 1800 P-I curves. These are then combined with satellite-derived chlorophyll data for the N. Atlantic, and information on cloud cover, to compute primary production at the annual scale. using a model that computes spectral transmission of light underwater, and spectral, photosynthetic response of phytoplankton to available light. The results are compared with earlier, satellite-derived, estimates of basin-scale primary production.

Ocean primary production and available light: further algorithms for remote sensing

Abstract In the context of remote sensing of the ocean, the general problem of estimating water column production from surface irradiance and column chlorophyll concentration is examined, and some refinements are made to the linear theory presented by Platt (1986, Deep-Sea Research , 33 , 149–163). Further empirical evidence is presented to show the stability of the relationship between surface light and biomass-normalized primary production of the ocean water column. A theoretical explanation is given for the non-zero intercept often obtained when these two variables are regressed. The systematic errors in the estimation of primary production by remote sensing, due to non-uniformity in the biomass profile, are examined through sensitivity analyses on a generalized biomass profile. The errors are shown to be functions of the parameters of the biomass profile, of the photosynthetic parameters and of the optical properties of the water. The probable random errors in the estimation of water column primary production using remotely sensed data are evaluated. Some general issues related to the collection and assimilation of data on ocean primary production are addressed.

Basin‐scale estimates of oceanic primary production by remote sensing: The North Atlantic

The estimation by remote sensing of annual primary production at ocean basin scales is illustrated for the Atlantic Ocean, using the monthly averaged Coastal Zone Color Scanner data for 1979. The principal supplementary data used were some 873 vertical profiles of chlorophyll and some 248 sets of parameters derived from photosynthesis-light experiments. This information was used to parametrize the local algorithm for calculation of primary production in 12 subregions of the entire domain for each of the four seasons. Four different procedures were tested for calculation of primary production. These differed according to whether the autotrophic biomass distribution was uniform with depth and whether the irradiance was resolved with respect to wavelength: the spectral model with nonuniform biomass was considered as the benchmark for comparison against the other three models. At particular locations and times, the less complete models gave results that differed by as much as 50% from the benchmark. After integration to basin scale, vertically uniform models tended to underestimate primary production by about 20% compared to the nonuniform models. At large horizontal scale, the differences between spectral and nonspectral models were negligible, a result that was believed to follow from mutual compensation of underestimates and overestimates, according to the local biomass, in different parts of the domain. Calculation of primary production is highly sensitive to the algorithm used to retrieve the biomass. The linear correlation between biomass and estimated production was poor outside the tropics, suggesting caution against the indiscriminate use of biomass as a proxy variable for primary production. The annual primary production for the Atlantic between 20°S and 70°N was 9 ± 3 Gt yr−1, higher than previous estimates made without reference to remotely sensed data. It is argued that the remote-sensing method is the method of choice for calculation of primary production at the ocean basin scale.

Remote sensing of oceanic primary production: computations using a spectral model

Abstract A spectral model of underwater irradiance is coupled with a spectral version of the photosynthesis-light relationship to compute oceanic primary production. The results are shown to be significantly different from those obtained using the conventional non-spectral approach. The problem of non-uniform vertical distribution of biomass is investigated next, from the point of view of estimation of water-column primary production using satellite data. The errors in estimated production are shown to be functions of parameters of the biomass distribution; of the photosynthesis parameters; and of the optical properties of the water column. Some examples are given to illustrate the comparison of model results with thee observed data.