A curated UK rain radar data set for training and benchmarking nowcasting models

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Atureta, Viv, Jasin, Rifki Priansyah, Siegert, Stefan
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908724592902144
author Atureta, Viv
Jasin, Rifki Priansyah
Siegert, Stefan
author_facet Atureta, Viv
Jasin, Rifki Priansyah
Siegert, Stefan
contents This paper documents a data set of UK rain radar image sequences for use in statistical modeling and machine learning methods for nowcasting. The main dataset contains 1,000 randomly sampled sequences of length 20 steps (15-minute increments) of 2D radar intensity fields of dimension 40x40 (at 5km spatial resolution). Spatially stratified sampling ensures spatial homogeneity despite removal of clear-sky cases by threshold-based truncation. For each radar sequence, additional atmospheric and geographic features are made available, including date, location, mean elevation, mean wind direction and speed and prevailing storm type. New R functions to extract data from the binary "Nimrod" radar data format are provided. A case study is presented to train and evaluate a simple convolutional neural network for radar nowcasting, including self-contained R code.
format Preprint
id arxiv_https___arxiv_org_abs_2512_17924
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A curated UK rain radar data set for training and benchmarking nowcasting models
Atureta, Viv
Jasin, Rifki Priansyah
Siegert, Stefan
Atmospheric and Oceanic Physics
Computer Vision and Pattern Recognition
Machine Learning
Applications
This paper documents a data set of UK rain radar image sequences for use in statistical modeling and machine learning methods for nowcasting. The main dataset contains 1,000 randomly sampled sequences of length 20 steps (15-minute increments) of 2D radar intensity fields of dimension 40x40 (at 5km spatial resolution). Spatially stratified sampling ensures spatial homogeneity despite removal of clear-sky cases by threshold-based truncation. For each radar sequence, additional atmospheric and geographic features are made available, including date, location, mean elevation, mean wind direction and speed and prevailing storm type. New R functions to extract data from the binary "Nimrod" radar data format are provided. A case study is presented to train and evaluate a simple convolutional neural network for radar nowcasting, including self-contained R code.
title A curated UK rain radar data set for training and benchmarking nowcasting models
topic Atmospheric and Oceanic Physics
Computer Vision and Pattern Recognition
Machine Learning
Applications
url https://arxiv.org/abs/2512.17924