Published June 17, 2024 | Version v2.0

Scripts for Paper Titled "Improving low-cloud fraction prediction through machine learning"

  • 1. ROR icon University of Maryland, College Park
  • 2. ROR icon Earth System Science Interdisciplinary Center
  • 3. ROR icon University of Houston

Description

Two zip files of scripts were attached, used in the AGU GRL paper titled "Improving low-cloud fraction prediction through machine learning". One file was used to set up the nudging simulations of CAM5 and CAM6, which produced 6-hourly output. The 6-hourly output was not included, as it totaled ~2.6TB. Another file was used to optimize and train two machine learning models used in the study, including XGB10 and XGB7. It was noted that Bayesian optimization was adopted for hyperparameter tuning.

Files

ML_codes.zip

Files (13.7 kB)

Name Size Download all
md5:48b169aead70219a57c9a0483668018c
9.8 kB Preview Download
md5:2a1de5e9f7c593271a922062a41910a1
3.9 kB Preview Download