Date of Award

2025

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Atmospheric and Earth Science

Committee Chair

Udaysankar Nair

Committee Member

Sundar Christopher

Committee Member

Aaron Kaulfus

Research Advisor

Udaysankar Nair

Subject(s)

Energy budget (Geophysics), Solar radiation, Meteorology, Artificial intelligence

Abstract

Estimating heat and moisture exchange between the land and atmosphere has several important practical applications, including water resource management, air pollution forecasting, and atmospheric propagation modeling. Turnkey systems for measuring the surface energy budget typically cost between $5,000 and over $50,000. This study explores the use of artificial intelligence (AI) to predict sensible heat flux (SHF) and latent heat flux (LHF) using inexpensive surface meteorology data and downwelling solar radiation measurements as inputs. Observations from Amer- iFlux sites were used to train several AI models—Convolutional Neural Networks (CNNs), One Dimensional Transformers (1D Transformers), Artificial Neural Net- works (ANNs), and Long Short-Term Memory (LSTM) networks. Among these, the 1D Transformer model demonstrated the best performance, with an average root mean squared error (RMSE) of 46 W/m² and a correlation coefficient of 0.88 for SHF, and an RMSE of 48 W/m² with a correlation of 0.85 for LHF, establishing the feasibility of using AI to predict components of the surface energy budget using meteorological data and downwelling solar radiation as predictors.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.