Hyperspectral Benchmark Dataset on Soil Moisture Hyperspectral Benchmark Dataset on Soil Moisture

* This dataset supports online use.Click here to jump.

Hyperspectral benchmark dataset on soil moisture is a soil moisture assessment benchmark dataset based on hyperspectral data. The dataset was obtained through a five-day field measurement campaign in Karlsruhe, Germany, focusing on undisturbed soil samples taken from the area near Waldbronn, Germany, without any vegetation, in the form of bare soil.

During this field campaign, the following sensors were used for measurements:

• Cubert UHD 285 hyperspectral snapshot camera, records 50 × 50 pixel images covering 125 spectral bands ranging from 450 nm to 950 nm with a spectral resolution of 4 nm.
• TRIME-PICO Time Domain Reflectometry (TDR) sensor that measures soil moisture percentage at a depth of 2 cm.

The variables in the dataset are:

• datetime:  Date and time of the measurement.
• soil_moisture:  Soil moisture percentage.
• soil_temperature:  Soil temperature, degrees Celsius.
• Spectral bands:  Hyperspectral band from 454 nm to 950 nm with 1 nm resolution.

The data set was measured and processed in order to study and develop models capable of estimating soil moisture content based on hyperspectral data.Introducing a Framework of Self-Organizing Maps for Regression of Soil Moisture with Hyperspectral Data" introduces this dataset and shows how to use the self-organizing map framework to perform soil moisture regression analysis. The paper has been accepted by IEEE 2018.