Dissertation · 2026 · Open access
Using generative AI to predict fine-scale weather and climate changes over New Zealand
UNSWorks (University of New South Wales, Sydney, Australia) · Published 13 February 2026 · DOI 10.26190/unsworks/32016
In brief
“Using generative AI to predict fine-scale weather and climate changes over New Zealand” is a 2026 dissertation by Neelesh Rampal, published in UNSWorks (University of New South Wales, Sydney, Australia). OpenAlex counts 0 citations as of 21 November 2026. It is open access under a CC BY licence: the full text is free to read at UNSWorks (University of New South Wales, Sydney, Australia).
- Citations
- 0
- Authors
- 1
Abstract
Global Climate Models (GCMs) simulate past, present, and future climates, but their coarse resolution limits their ability to resolve fine-scale processes. While Regional Climate Models (RCMs) address this issue, their high computational cost restricts the number of GCMs that can be downscaled. Internal variability is an important yet under-studied source of uncertainty at fine spatial scales and reliably separating it from the forced climate response would require larger downscaled ensembles than is currently feasible. Artificial intelligence (AI) offers a computationally efficient way to emulate RCMs and rapidly downscale large GCM ensembles. The primary objective of this thesis is to investigate whether AI-based approaches can reliably emulate RCMs and subsequently be used to separate internal variability from forced climate responses in climate projections at fine spatial scales. Chapter 2 synthesizes the literature on AI-based downscaling and identifies two important limitations: AI methods often fail to predict extremes at fine spatial scales (Chapter 3) and struggle to extrapolate to unobserved, warmer climates (Chapter 4). Chapter 3 develops a generative adversarial network (GAN) approach to emulate an RCM, which accurately predicts fine-scale historical climatologies of mean and extreme precipitation as well as day-to-day variability. However, the analysis shows that incorporating constraints into the GAN is important, as it improves its skill and robustness. Chapter 4 evaluates how the GAN extrapolates beyond its training distribution by comparing predictions against out-of-sample RCM simulations. This reveals that training on warmer future simulations, rather than historical ones, is important for capturing future climate change responses of mean and extreme precipitation. Chapter 5 implements the above methodological developments and applies the GAN to downscale over 15,000 years of GCM simulations, producing a large ensemble of high-resolution climate projections to investigate internal variability in rare climate extremes at fine spatial scales. Results show that internal variability often exceeds the forced response for precipitation extremes and is the dominant source of projection uncertainty. Furthermore, the analysis shows that internal variability is larger at fine scales than at coarser GCM scales, suggesting GCMs may underestimate this variability. Overall, fine-scale changes in precipitation extremes are less predictable than often assumed and require large, downscaled ensembles to assess reliably.
Cite this paper
Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.
APA 7
Rampal, N. (2026). Using generative AI to predict fine-scale weather and climate changes over New Zealand. UNSWorks (University of New South Wales, Sydney, Australia). https://doi.org/10.26190/unsworks/32016
MLA 9
Rampal, Neelesh. “Using generative AI to predict fine-scale weather and climate changes over New Zealand.” UNSWorks (University of New South Wales, Sydney, Australia), 2026. https://doi.org/10.26190/unsworks/32016.
Chicago
Rampal, Neelesh. 2026. “Using generative AI to predict fine-scale weather and climate changes over New Zealand.” UNSWorks (University of New South Wales, Sydney, Australia). https://doi.org/10.26190/unsworks/32016.
Harvard
Rampal, N. (2026) ‘Using generative AI to predict fine-scale weather and climate changes over New Zealand’, UNSWorks (University of New South Wales, Sydney, Australia). doi:10.26190/unsworks/32016.
IEEE
N. Rampal, “Using generative AI to predict fine-scale weather and climate changes over New Zealand,” UNSWorks (University of New South Wales, Sydney, Australia), 2026, doi: 10.26190/unsworks/32016.
BibTeX
@phdthesis{rampal2026using,
title = {{Using generative AI to predict fine-scale weather and climate changes over New Zealand}},
author = {Rampal, Neelesh},
year = {2026},
doi = {10.26190/unsworks/32016},
url = {https://doi.org/10.26190/unsworks/32016}
}
Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)
Topic and keywords
Physical Sciences › Environmental Science › Global and Planetary Change › Climate variability and models
- generative adversarial networks
- regional climate model
- global climate model ensembles
- natural climate variability
- heavy rainfall
- uncertainty in climate projections
- New Zealand climate
Work with this paper
Most-cited papers in Environmental Science
- Biodiversity hotspots for conservation prioritiesNorman Myers, Russell A. Mittermeier, Cristina Goettsch Mittermeier and 2 more · Nature · 2000 · 32,720 citations
- The Tragedy of the CommonsGarrett Hardin · Science · 1968 · 26,194 citations
- River flow forecasting through conceptual models part I — A discussion of principlesJ.E. Nash, John V. Sutcliffe · Journal of Hydrology · 1970 · 24,625 citations
- The value of the world's ecosystem services and natural capitalRobert Costanza, Ralph C. d’Arge, R.S. de Groot and 10 more · Nature · 1997 · 23,508 citations
- Climate change 2007: the physical science basisSusan Solomon, Dahe Qin, Martin Manning and 5 more · 2007 · 23,383 citations
Questions about this paper
How do I cite this paper in APA style?
Rampal, N. (2026). Using generative AI to predict fine-scale weather and climate changes over New Zealand. UNSWorks (University of New South Wales, Sydney, Australia). https://doi.org/10.26190/unsworks/32016
How many times has this paper been cited?
0 times, according to OpenAlex on 21 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.
Is this paper free to read?
Yes. It is open access at UNSWorks (University of New South Wales, Sydney, Australia): sign in and open the full text from this page.
Who wrote this paper?
Neelesh Rampal.
What is the DOI of this paper?
10.26190/unsworks/32016, permanently at https://doi.org/10.26190/unsworks/32016. Sign in and open it from this page.
Bibliographic data from OpenAlex (CC0), last checked 21 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.