Late Pleistocene–Holocene radiolarian paleotemperatures in the Norwegian Sea based on artificial neural networks

Abstract

Artificial Neural Networks (ANN) were trained by using an extensive radiolarian census dataset from the Nordic (Greenland, Norwegian, and Iceland) Seas. The regressions between observed and predicted Summer Sea Temperature (SST) indicate that lower error margins and better correlation coefficients are obtained for 100 m (SST 100 ) compared to 10 m (SST 10 ) water depth, and by using a subset of species instead of all species. The trained ANNs were subsequently applied to radiolarian data from two Norwegian Sea cores, HM 79-4 and MD95-2011, for reconstructions of SSTs through the last 15,000 years. The reconstructed SST is quite high during the Bølling-Allerød, when it reaches values only found later during the warmest phase of the Holocene. The climatic transitions in and out of the Younger Dryas are very rapid and involve a change in SST 100 of 6.2 and 6.8 8C, taking place over 440 and 140 years, respectively. SST 100 remains at a maximum during the early Holocene, and this Radiolarian Holocene Optimum Temperature Interval (RHOTI) predates the commonly recognized middle Holocene Climatic Optimum (HCO). During the 8.2 ka event, SST 100 decreases by ca. 3 8C, and this episode marks the establishment of a cooling trend, roughly spanning the middle Holocene (until ca. 4.2 ka). Successively, since then and through the late Holocene, SST 100 follows instead a statistically significant warming trend. The general patterns of the reconstructed SSTs agree quite well with previously obtained results based on application of Imbrie and Kipp Transfer Functions (IKTF) to the same two cores for SST 0 . A statistically significant cyclic component of our SST record (period of 278 years) has been recognized. This is close to the de Vries or Suess cycle, linked to solar variability, and documented in a variety of other high-resolution Holocene records. D

The Holocene thermal maximum in the Nordic Seas: the impact of Greenland Ice Sheet melt and other forcings in a coupled atmosphere–sea-ice–ocean model

“…between 8.5 and 5.5 ka BP) at the Vøring Plateau. Moreover, at the same site, SST reconstructions based on foraminifera and radiolarians also lack this early HTM suggested by diatoms (Risebrobakken et al, 2003(Risebrobakken et al, , 2011Cortese et al, 2005;Dolven et al, 2002). Jansen et al (2008) suggest that the SST maximum recorded in proxy data above the seasonal thermocline (diatoms, alkenones) is forced by the summer insolation maximum and that deeper dwelling species (foraminifera and radiolarians) are not influenced.…”

Anomalous radioactivity level and high concentrations of heavy minerals in Lemme area, South-West Estonia

“…The most accurate data were produced by the NGO. Cortese et al (2005) found, based on three test runs, that the NGO generated slightly lower error rates than the iModel.…”

“…The network output is compared to the assumed output, which is specified by the computer simulation (as supervisor), and the error for all the neurons in the output layer is calculated. The essential mechanism behind back propagation is that the error is propagated backward to earlier layers so that a gradient descent algorithm can be applied (Malmgren, Nordlund 1997;Malmgren et al 2001;Cortese et al 2005;Kucera et al 2005).…”

“…This software provided the best results from ANN tests conducted in a study by Cortese et al (2005). They tested by means of a radiolarian reference data set and summer sea surface temperature the two different ANN programs: the NeuroGenetic Optimizer (NGO; ©BioComp Systems, Inc.) and the iModel (©BioComp Systems, Inc.).…”

“…They therefore provide estimates of the prediction error rate in samples different than those used in the training set (Cortese et al 2005). Thus, obtaining an RMSE of 0.5, if the estimated salinity is 8 ‰, means that the estimated error in palaeosalinity data is at least 0.5 ‰.…”