Agrineer - Campaign for soil moisture sensor comparison

This is a (future) crowd-funding campaign to study how in-house open-source soil moisture sensors and data acquisition units compare to proprietary research grade soil moisture sensors and data acquisition units.
Motivation

Agrineer.org's SME project strives to give an accurate estimate of available soil moisture for a crop field through a series of numerical modeling steps. This campaign is the first part of an effort to calibrate the SME.

Many researchers write excellent papers on the mechanics of agricultural evaporation but fail to implement their methods on a large scale due many to reasons; complexity, local data sensitivities, lack of funding, and follow-through to mention a few.

Approach

Our approach is to reduce the complexities enough where large scale environmental conditions can be calculated and adjusted using local conditions (ie. cultivars, soil, and precipitation events), with the intent of providing useful information to the farmer, based on their environment. Standard evapotranspiration (ETo) values from the SME initial release give reasonable starting values to work with in order to improve the accuracy of the SME.

Tools

Static soil moisture sensors are the preferred method to "ground truth" the calculated values given by the SME application. These sensors are connected to a data collector and sampled periodically.

There are two commonly used methods to sense the moisture using the electrical properties of the soil, one is by measuring the capacitance and the other is to send a signal through the soil and analyzing the waveform reflection. The latter method is a type of radar and is called a time-domain reflectometer (TDR).

The TDR sensors are more accurate and more expensive than the capacitance sensors. Current research papers have been written describing good results for a mixed network of sensors using both types of sensors, where numerous low-cost sensors are calibrated using the more expensive TDR. These papers indicate that it is better to have large spatial coverage with numerous calibrated low-cost sensors than to have just a few high-cost sensors.

We are currently developing/integrating open-source versions of the data collector and sensors, using Arduino, Rasberries, Software Defined Radio (SDR), and other components. Once deemed fit for duty, these sensors will be compared to proprietary research grade units and then to the SME calculated values.

The approach is split into two parts:

  • compare in-house open-source soil moisture sensors to research-grade sensors (this campaign).
  • compare soil moisture sensors to SME values (future campaign).
Campaign Goal

Our goal is to raise sufficient funding to:

  • purchase research-grade sensor hardware
  • conduct field data acquisition over the growing and non-growing seasons using in-house and research-grade sensors
  • analyse and report.
A detailed breakout will be available at the crowd-funding site (to be determined). Data and results will be submitted for publication.

Please check back later to see when we have launched this campaign.