Microwave and millimeter wave signals reflectance of soil carbon content

An, Di1 ; Chen, YangQuan1

Research facility: University of California, Merced

Published May 24, 2023; Updated Apr 15, 2024 on Dryad. https://doi.org/10.6071/M3M092

Abstract

Agricultural and forestry biomass can be converted to biochar through pyrolysis gasification, making it a significant carbon source for soil. Applying biochar to soil is a carbon-negative process that helps combat climate change, sustain soil biodiversity, and regulate water cycling. However, quantifying soil carbon content conventionally is time-consuming, labor-intensive, imprecise, and expensive, making it difficult to accurately measure in-field soil carbon's effect on storage water and nutrients. To address this challenge, for the first time, we report on extensive lab tests demonstrating non-intrusive methods for sensing soil carbon and related smart biochar applications, such as differentiating between biochar types from various biomass feedstock species, monitoring soil moisture, and biochar water retention capacity using portable microwave and millimeter wave sensors and machine learning. The datasets provide details on the microwave and millimeter wave reflectance signals. We validated our quantification method using supervised machine learning algorithms by collecting real soil mixed with known biochar contents in the field.

README: Microwave and Millimeter wave signals reflectance of Soil Carbon Content

Reference Information

Authors: Di An, YangQuan Chen

Data Created: 04/10/2023

Data Modified: 05/08/2023

Suggested Citations: Dataset: An, Di; Chen, YangQuan (2023), Microwave and millimeter wave signals reflectance of soil carbon content, Dryad, Dataset, https://doi.org/10.6071/M3M092

Corresponding publication: An, D., & Chen, Y. (2023). Non-intrusive soil carbon content quantification methods using machine learning algorithms: A comparison of microwave and millimeter wave radar sensors. Journal of Automation and Intelligence, 2(3), 152-166.

Data and File Overview

There are a total of 10 datasets: 5 for microwave and 5 for millimeterwave, respectively. The name meaning as follows:
"amount_biochar_microwave", "final_soilContent_low_millimeterwave" = Different Soil Carbon Amount Recognition
"type_biochar_microwave", "final_biochar_low_millimeterwave" = Different Types of Soil Carbon Classification
"soil_moisture_microwave", "final_moisture_low_millimeterwave" = Soil Moisture Study and Comparison
"Mos_C_microwave", "final_moistureCC_low_millimeterwave" = Soil Moisture Influenced by Carbon Content Added
"multiple_containers_microwave", "millimeterwave_mul_containers" = Sensing Robustness for Soil Carbon Content Percentage

Setup

Unpacking instructions: n/a
Relationships between files/folders: they are in different experiments sections
Recommended software/tools: > Matlab 2020

File/Folder Details-Microwave

Details for: amount_biochar_microwave.mat

Description: This is for classifying different amount of biochar mixed with soil using microwave sensor.
Format(s): .mat
Size(s): 6,149 KB
Dimensions:
- Data: 50 rows x 8192 cols
- label: 50 rows x 1 cols
Variables:
- amount_biochar:
  - type: table
  - description: combined sampling data with label together. Can be used in Classifier App directly.
- d_2: raw data
- y: Total 5 classes, each label has 10 sampling, therefore totally it has 50 labels.

Details for: type_biochar_microwave.mat

Description: This is for classifying different types of biochar using microwave sensor.
Format(s): .mat
Size(s): 6,140 KB
Dimensions:
- Data: 50 rows x 8192 cols
- label: 50 rows x 1 cols
Variables:
- type_biochar:
  - type: table
  - description: combined sampling data with label together. Can be used in Classifier App directly.
- d_2: raw data
- y: Total 5 classes, each label has 10 sampling, therefore totally it has 50 labels.

Details for: Mos_C_microwave.mat

Description: This is for classifying different soil moisture affected by carbon content added
Format(s): .mat
Size(s): 6,150 KB
Dimensions:
- Data: 50 rows x 8192 cols
- label: 50 rows x 1 cols
Variables:
- mos_carbon:
  - type: table
  - description: combined sampling data with label together. Can be used in Classifier App directly.
- d_2: raw data
- y: Total 5 classes, each label has 10 sampling, therefore totally it has 50 labels.

Details for: multiple_contains_microwave.mat

Description: This is for investigating robustness sensing for soil carbon content percentage
Format(s): .mat
Size(s): 61,322 KB
Dimensions:
- Data: 500 rows x 8192 cols
- label: 500 rows x 1 cols
Variables:
- mulitle_container:
  - type: table
  - description: combined sampling data with label together. Can be used in Classifier App directly.
- d_2: raw data
- y: Total 5 classes, each label has 10 sampling, therefore totally it has 50 labels.

Details for: soil_moisture_microwave.mat

Description: This is for classifying different soil moisture levels using microwave sensor.
Format(s): .mat
Size(s): 6,148 KB
Dimensions:
- Data: 50 rows x 8192 cols
- label: 50 rows x 1 cols
Variables:
- moisture_soil:
  - type: table
  - description: combined sampling data with label together. Can be used in Classifier App directly.
- d_2: raw data
- y: Total 5 classes, each label has 10 sampling, therefore totally it has 50 labels.

File/Folder Details-Millimeter wave

Details for: final_soilContent_low_millimeterwave.mat

Description: This is for classifying different amount of biochar mixed with soil using millimeter wave sensor.
Format(s): .mat
Size(s): 2,134 KB
Dimensions:
- Data: 140 rows x 1024 cols
- label: 140 rows x 1 cols
Variables:
- soilCarbon_low:
  - type: table
  - description: combined sampling data with label together. Can be used in Classifier App directly.
- new_soilContent_low: raw data
- y: Total 7 classes, each label has 20 sampling, therefore totally it has 140 labels.

Details for: final_biochar_low_millimeterwave.mat

Description: This is for classifying different types of biochar using millimeter wave sensor.
Format(s): .mat
Size(s): 1,857 KB
Dimensions:
- Data: 120 rows x 1024 cols
- label: 120 rows x 1 cols
Variables:
- newBiochar:
  - type: table
  - description: combined sampling data with label together. Can be used in Classifier App directly.
- new_biochar_low: raw data
- y: Total 6 classes, each label has 20 sampling, therefore totally it has 120 labels.

Details for: final_moistureCC_low_millimeterwave.mat

Description: This is for classifying different soil moisture affected by carbon content added using millimeter wave sensor.
Format(s): .mat
Size(s): 2,475 KB
Dimensions:
- Data: 160 rows x 1024 cols
- label: 160 rows x 1 cols
Variables:
- new_Mosi:
  - type: table
  - description: combined sampling data with label together. Can be used in Classifier App directly.
- new_moistureC_low: raw data
- y: Total 8 classes, each label has 20 sampling, therefore totally it has 160 labels.

Details for: mm_multi_containers.mat

Description: This is for investigating robustness sensing for soil carbon content percentage using millimeter wave sensor.
Format(s): .mat
Size(s): 3,279 KB
Dimensions:
- Data: 500 rows x 300 cols
- label: 500 rows x 1 cols
Variables:
- mm_mult_container:
  - type: table
  - description: combined sampling data with label together. Can be used in Classifier App directly.
- mul_contain_mm: raw data in double types, can be used directly.
- multi_containers_mm: raw data in complex number, easy for doing PDP plots.
- y: Total 5 classes, each label has 100 sampling, therefore totally it has 500 labels.

Details for: final_moisture_low_millimeterwave.mat

Description: This is for classifying different soil moisture levels using microwave sensor.
Format(s): .mat
Size(s): 2,166 KB
Dimensions:
- Data: 140 rows x 8192 cols
- label: 140 rows x 1 cols
Variables:
- new_Mosi:
  - type: table
  - description: combined sampling data with label together. Can be used in Classifier App directly.
- new_moisture_low: raw data
- y: Total 7 classes, each label has 20 sampling, therefore totally it has 140 labels.