Daily light integral distributions: geographic similarity

Every crop species has a preferred daily light integral (DLI) for optimal plant health and growth.
For field crops, maps showing the monthly DLI in various regions can be prepared from geographic locations and historical weather data.
However, the original method depends on historical weather data from weather stations located in the continental United States.
While similar data are available from weather stations worldwide, the stations are sparsely distributed, making DLI distribution calculations unreliable.
An updated method uses a solar radiation model, but it relies on satellite data that are only available for the United States.
We therefore propose a new method that combines global weather station data with solar radiation data obtained from meteorological satellites to provide worldwide coverage, and also a neural network approach that assigns a geographic similarity metric to weather stations that enables DLI distributions to be calculated using distant but geographically similar weather station data.

IX International Symposium on Light in Horticulture

J. Thomas, I. Ashdown

DLI distribution maps, TMY weather records, solar energy models

https://doi.org/10.17660/ActaHortic.2022.1337.35