Parameter estimation for land-surface models using Neural Physics

Huang, Ruiyue; Heaney, Claire E.; van Reeuwijk, Maarten

doi:10.5194/gmd-19-8801-2026

Articles | Volume 19, issue 18

https://doi.org/10.5194/gmd-19-8801-2026

© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/gmd-19-8801-2026

© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 19, issue 18

Development and technical paper

|

21 Sep 2026

Development and technical paper |

| 21 Sep 2026

Parameter estimation for land-surface models using Neural Physics

Ruiyue Huang, Claire E. Heaney, and Maarten van Reeuwijk

Data sets

Eddy covariance data measured at the CAP LTER flux tower located in the west Phoenix, AZ neighborhood of Maryvale from 2011-12-16 through 2012-12-31 ver 1 W. Chow https://doi.org/10.6073/pasta/fed17d67583eda16c439216ca40b0669

Data for "Harmonized gap-filled dataset from 20 urban flux tower sites" for the Urban-PLUMBER project M. Lipson et al. https://doi.org/10.5281/zenodo.7104984

Model code and software

Solving land surface model using Neural Physics R. Huang https://doi.org/10.5281/zenodo.19344692

Short summary

This paper uses the Neural Physics approach to determine parameters of a simple land-surface model. We show that we can only obtain a reliable parameter estimation using soil temperature measurements at more than one depth, and that latent and sensible heat fluxes cannot be differentiated. We then apply the inverse model to real urban flux tower data and show that parameters, as well as various heat fluxes, can be reliably estimated using an observed value for the effective surface albedo.