3 research outputs found
Novel deep learning algorithm in soil erodibility factor predicting at a continental scale
No full textSoil erosion poses significant environmental and economic challenges, adversely affecting soil fertility and global agricultural productivity. We developed a novel model based on the Multi-Head Squeeze-and-Excitation Residual One-Dimensional Convolutional Neural Network (MH-SE-Res1DNet) to predict the soil erodibility factor (K) across Europe, representing the first application of this model for such a purpose worldwide. We conducted a comparative analysis using five benchmark machine learning algorithms, i.e., Random Forest (RF), Artificial Neural Network–Multilayer Perceptron (ANN-MLP), Support Vector Regression (SVR), Alternating Model Tree (AMT), and Pace Regression (PR), to assess the efficacy of our model. The results showed that the MH-SE-Res1DNet deep learning model had an outstanding ability for the K prediction. The model's lowest error (MAE = 0.0025, RMSE = 0.0031) and highest coefficient of determination (R2 = 0.943) were attained during the validation phase. Benchmark models demonstrated lower performance compared to the MH-SE-Res1DNet model, with R2 values ranging from 0.880 to 0.912 and slightly higher errors across MAE and RMSE metrics. The sensitivity analysis of MH-SE-Res1DNet showed that its performance depends predominantly on key soil factors, particularly topsoil texture (M) and organic matter (OM) concentration. This model establishes a data-driven framework that significantly advances soil erodibility prediction by leveraging machine learning. It surpasses traditional methods and existing machine learning approaches in accuracy, efficiency, and scalability, setting a new benchmark for soil conservation planning and enabling adaptable, evidence-based land management strategies across Europe and worldwide.JRC.D.1 - Land and Climat
COHERENT SYNCHROTRON RADIATION FOR ROTATIONAL SPECTROSCOPY: APPLICATION TO THE ROTATIONAL SPECTRUM OF PROPYNAL IN THE 200-750 GHz RANGE
No full textAuthor Institution: Synchrotron SOLEIL, L'Orme des Merisiers Saint-Aubin, 91192 Gif-sur-Yvette, France; Australian Synchrotron Light Source, 800 Blackburn Road Victroria 3168 and Monash University, School of Chemistry, Clayton, Victoria 3800 Australia.; La Trobe University, Department of Chemistry, Victoria, 3086, Australia; Laboratoire LADIR, CNRS, Universite Pierre et Marie Curie, 75252 Paris Cedex, FranceIn storage rings, short electron bunches can produce an intense THz radiation called Coherent Synchrotron Radiation (CSR). The flux of this emission between 250 and 750 GHz (in the mW range, up the 10000 times the regular synchrotron emission) is very advantageous for broad band absorption spectroscopy, using interferometric techniques. This source is, however, inherently difficult to stabilize, and intensity fluctuations lead to artifacts on the FT-based measurements, which strongly limit the use of CSR in particular for high-resolution measurements. At SOLEIL however, by screening different currents and bunch lengths, we defined stable CSR conditions for which the signal-to-noise ratio (S/N) allows for measurements at high resolution. Moreover, we developed an artifact correction system, based on a simultaneous detection of the input and the output signals of the interferometer, which allows to further improve the S/N. For this purpose, the optics and electronics of two bolometers were matched. The stable CSR combined with this ingenious technique allowed us to record for the first time high-resolution FT spectra in the sub-THz range, with a S/N of 100 in a few hours. This enables many applications such as broadband rotational spectra in the THz range, studies of molecules with low frequency torsional modes, absolute intensities determinations, or studies of unstable species. Results obtained on Propynal illustrate these possibilities and enabled to improve significantly the ground state spectroscopic constants
Causes of decoupling between larval supply and settlement and consequences for understanding recruitment and population connectivity
No full textAuthor Posting. © The Author(s), 2010. This is the author's version of the work. It is posted here by permission of Elsevier B.V. for personal use, not for redistribution. The definitive version was published in Journal of Experimental Marine Biology and Ecology 392 (2010): 9-21, doi:10.1016/j.jembe.2010.04.008.Marine broadcast spawners have two-phase life cycles, with pelagic larvae and benthic adults.
Larval supply and settlement link these two phases and are crucial for the persistence of marine
populations. Mainly due to the complexity in sampling larval supply accurately, many
researchers use settlement as a proxy for larval supply. Larval supply is a constraining variable
for settlement because, without larval supply, there is no settlement. Larval supply and
settlement may not be well correlated, however, and settlement may not consistently estimate
larval supply.
This paper explores the argument that larval supply (i.e., larval abundance near settlement sites)
may not relate linearly to settlement. We review the relationship between larval supply and
settlement, from estimates and biases in larval supply sampling, to non-behavioral and
behavioral components, including small-scale hydrodynamics, competency, gregarious behavior,
intensification of settlement, lunar periodicity, predation and cannibalism. Physical and structural
processes coupled with behavior, such as small-scale hydrodynamics and intensification of
settlement, sometimes result in under- or overestimation of larval supply, where it is predicted
from a linear relationship with settlement. Although settlement is a function of larval supply,
spatial and temporal processes interact with larval behavior to distort the relationship between
larval supply and settlement, and when these distortions act consistently in time and space, they
cause biased estimates of larval supply from settlement data.
Most of the examples discussed here suggest that behavior is the main source of the decoupling
between larval supply and settlement because larval behavior affects the vertical distribution of
larvae, the response of larvae to hydrodynamics, intensification of settlement, gregariousness,
predation and cannibalism. Thus, larval behavior seems to limit broad generalizations on the
regulation of settlement by larval supply. Knowledge of the relationship is further hindered by
the lack of a well founded theoretical relationship between the two variables.
The larval supply- settlement transition may have strong general consequences for population
connectivity, since larval supply is a result of larval transport, and settlement constrains
recruitment. Thus, measuring larval supply and settlement effectively allows more accurate
quantification and understanding of larval transport, recruitment and population connectivity.JP would like to thank WHOI Ocean Life Institute for partial funding. FP’s contribution is based upon research supported by the South African Research Chairs Initiative of the Department of Science and Technology and National Research Foundation
