API src

Found 5 results.

Other language confidence: 0.8171399653637383

Model files for the Neural network-based model of Electron density in the Topside ionosphere (NET)

Here, we present model files and example scripts for the Neural network-based model of Electron density in the Topside ionosphere (NET). The model is based on radio occultation data from Gravity Recovery And Climate Experiment (GRACE), Challenging Minisatellite Payload (CHAMP) and Constellation Observing System for Meteorology Ionosphere and Climate (COSMIC-1) missions from 2001 until 2019. The NET model is based on alpha-Chapman functions with a linear decay of scale height with altitude, and consists of 4 sub-models (2 parameters of the F2-peak and 2 parameters of the linear scale height decay). The model uses geographic and magnetic latitude and longitude, magnetic local time, day of year, altitude, solar flux index P10.7, geomagnetic activity index Kp, storm-time SYM-H index as inputs. An example data frame to run the model is provided, as well as the Jupyter notebook to perform an example run.

Deep neural network enhanced global tropospheric zenith delay model

With the growing use of airborne platforms in Earth observation, accurate tropospheric delay corrections across various altitudes have become essential. Most existing tropospheric delay models are referenced to the Earth’s surface and rely on analytical closed-form vertical adjustments to approximate delays at user heights. However, these analytical models often fail to capture the complex vertical variations in atmospheric conditions. To address this limitation, we developed a novel approach leveraging deep neural networks (DNN) to reconstruct global three-dimensional zenith hydrostatic delay (ZHD) and zenith wet delay (ZWD) derived from numerical weather models (NWM). Using the ERA5 pressure-level product (Hersbach et al., 2020) for model training, the DNN refines predictions by correcting the residuals of an analytical third-order exponential model (EXP3). This hybrid method takes advantage of the non-linear fitting capabilities of DNN, significantly enhancing the accuracy of vertical tropospheric delay corrections up to 14 km above the Earth’s surface. The model achieves an average precision of 0.4 mm for ZHD and 0.8 mm for ZWD, reducing root-mean-square (RMS) errors by 63% and 36%, respectively, compared to EXP3. This dataset includes the EXP3 model, structured on a 1° × 1° global grid at four synoptic times daily (00:00, 06:00, 12:00, and 18:00 UTC) for the period 2019–2022. Additionally, it provides the corresponding DNN model to correct errors in the EXP3 predictions. It is important to note that the model is designed for altitudes ranging from the Earth’s surface up to 14 km.

Electron density derived with the Neural-network-based Upper-hybrid Resonance Determination algorithm from the Van Allen Probes EMFISIS measurements

The dataset presents the electron density derived using the Neural-network-based Upper hybrid Resonance Determination (NURD) algorithm (Zhelavskaya et al., 2016) from plasma wave measurements made with the Van Allen Probes Electric and Magnetic Field Instrument Suite and Integrated Science (EMFISIS) (Kletzing et al., 2013). The method employs feedforward neural networks to derive the upper hybrid resonance frequency from the electric field measurements, and hence electron density, in an automated fashion. The dataset contains electron density for the period from October 1, 2012 to January 14, 2018 for RBSP-A and from October 1, 2012 to July 1, 2016 for RBSP-B (RBSP = Radiation Belt Storm Probes).For convenience, the density data are organized in two ways: in terms of orbits and in terms of days. Directories ../../Orbits_organization/ and ../../Days_organization/ contain files with densities per orbit and per day, respectively. Data are provided in .txt and .cdf formats. Data in .mat format are available at ftp://ftp.gfz-potsdam.de/home/rbm/NURD/. For more information on directory organization and files description, please refer to the associated data description and Zhelaskaya et al. (2016).

The Brazilian gravimetric geoid: MAPGEO2015

The official Brazilian geoid model, called MAPGEO2015, was computed as a cooperation between IBGE and EPUSP. It is a 5' x 5' grid of geoidal undulations referred to SIRGAS2000, covering the area from 35° S to 6° N in latitude and from 75° W to 30° W in longitude. The dataset consists of 947,953 terrestrial gravity points (450,589 in Brazil), a digital terrain model based on SRTM and the EIGEN-6C4 global geopotential model up to degree and order 200. The remove-compute-restore procedure was used, where the short wavelength components were computed by Stokes integration via Fast Fourier Transform technique. In areas where sufficient data were available, neural network technique was used to complete the gravity anomaly information. Through the values of normal-orthometric heights of 592 geometric levelling points (RRNN) of the Brazilian High Precision Altimetric Network, it was possible to evaluate the consistency of the geoid undulation model, observing an improvement of approximately 20% compared to the version published in 2010. In particular, the differences between MAPGEO2015 undulations and GNSS/levelling observations show an RMS of 17 cm. The geoid model is provided in ISG format 2.0 (ISG Format Specifications), while the file in its original data format is available at the model ISG webpage.

AIMS: Automated Identification and Characterisation of Microbial Populations

The projekts overall aim is to develop, test and apply analytical procedures for identifying and characterising phytoplankton and heterotrophic bacterial populations using analytical flow cytometry (AFC). By identification, we mean the use of objective procedures for differentiating amongst populations of phytoplankton and bacteria within complex natural microbial communities. This involves classifying a community of heterogeneous organisms into its component populations. Identification will be based on optical properties of single cells measured by AFC and verified using molecular biology. By characterisation, we mean the determination of cell abundance together with an analysis of the intrinsic optical and chemical properties of these cells. These intrinsic properties include cell size, cell light scattering cross-section, cell carbon content, and additionally for phytoplankton, the light absorption cross-section and cell chlorophyll a content. The output of the identification and characterisation procedures will be a data matrix summarising the intrinsic cell properties of objectively defined populations. The products of this research will have application in large scale initiatives such as remote sensing and modelling basin scale and global oceanographic processes.

1