A New Weather Station Project for G-Mac's Country
Feb 11, 2022
Written By: Bryden Templeton
It is nearly impossible to undervalue the influence of weather on dryland grain production. While it's far from the only influence on production, it often determines the ceiling of any acre's ability to produce for a particular growing season. To further complicate things, the weather continues to be impossible to manage. It remains very challenging to predict with certainty, and can vary wildly between seasons and between different areas. Even during my short career, I've played witness to record-breaking rainfalls in 2016 and the most recent record-breaking drought of 2021. All of this in a span of 5 years. The physchology of an individual who relies solely on the probability of specific weather conditions for their livlihood is an interesting one to say the least (of course, without mentioning the uncertainty that comes with participating in global commodity markets as well). It's no wonder almost every casual conversation amongst folks engaged in agriculture includes a weather update.
The Ag Innovation group at G-Mac's AgTeam has kept a keen eye on the developing area of remote weather sensors, digital tools for viewing and managing data, and decision support systems informed by captured weather data for four years now. We have learned a lot and will continue to do so. Of course, like any new technology or approach, we have discovered some challenges along the way. Weather conditions can often vary in just mere miles which necessitates a lot of hardware to capture those conditions. Publicly accessible stations exist but are often too far away to rely on accurate data. Interpolated models meant to guess weather conditions between publicly available stations have been very inaccurate, again, due to the distance between hardware. Today, deploying multiple cellular-enabled weather stations on an average-sized farm has remained cost prohibitive to many. Issues with station performance, maintenance, and sensor calibration can be all too common as well.
Not done yet. User interfaces for viewing and interacting with data can be cumbersome and frustrating. Even the ability of exporting data for use in Microsoft Excel is not necessarily a given. In situations where an Agronomist or farmer is working with multiple brands of hardware, data is likely fragmented across different digital tools and awkward to compile and compare. Actionable insights beyond immediate real-time readings are often untapped and can be abstract. In other words, while using the wind sensor to assess spray risk is certainly an immediate and worthwhile application, I'm also interested in, say, the relationships between nighttime temperatures and heat blasting in canola; hard to analyze without reliable, local temperature data at scale. We are seeing lots of work being done with turning basic sensor data into a generated recommendation for spray timing and disease risk. If accurate, models like this will add to the value of capturing weather data.
Raising a crop is a scientific endeavour. One of the foundations of scientific methodology relates to the practice of identifying, isolating, and controlling for variables. Rainfall, air temperature, and humidity are examples of variables and they are often missing pieces of the agronomy puzzle. Motivated by the potential insights into grain production and being cognizant of the challenges outlined above, G-Mac's AgTeam is piloting a strategic weather sensor infrastructure project unlike anything attempted previously. The project will be an evaluation of benefit to an installed grid of weather stations that aims to strike a balance between resolution and investment of hardware. Next spring, G-Mac's AgTeam will select three branches to install 50-60 weather stations to capture the conditions for those geographies. Customers will have the option to access the weather data under a subscription model. This approach will increase the accessibility for many more farmers interested in a local weather network without the upfront cost in hardware and responsibility of installation and maintenance. In addition to the weather data, G-Mac's AgTeam has partnered with a company whose expertise in predictive analytics has enabled them to build models for crop staging and assessing disease risk for sclerotinia and fusarium head blight. This information will be available remotely in a mobile app.
This is a new and exciting project being led by our Ag Innovation team. Fortunately, G-Mac's AgTeam has the resources, scale, and foresight to explore innovative initiatives such as this. As agriculture continues down a path of realizing previously unavailable insights derived from data analytics at scale, the capturing and digitizing of major agronomic data point fundamental to grain production and will only become more necessary (weather conditions, soil attributes, input management, yield, etc). There is no example of an agricultural weather sensor network at this scale and resolution in Western Canada. If this project shows promise in terms of value, interest, and execution, G-Mac's AgTeam will be further exploring onboarding more locations.
It is nearly impossible to undervalue the influence of weather on dryland grain production. While it's far from the only influence on production, it often determines the ceiling of any acre's ability to produce for a particular growing season. To further complicate things, the weather continues to be impossible to manage. It remains very challenging to predict with certainty, and can vary wildly between seasons and between different areas. Even during my short career, I've played witness to record-breaking rainfalls in 2016 and the most recent record-breaking drought of 2021. All of this in a span of 5 years. The physchology of an individual who relies solely on the probability of specific weather conditions for their livlihood is an interesting one to say the least (of course, without mentioning the uncertainty that comes with participating in global commodity markets as well). It's no wonder almost every casual conversation amongst folks engaged in agriculture includes a weather update.
The Ag Innovation group at G-Mac's AgTeam has kept a keen eye on the developing area of remote weather sensors, digital tools for viewing and managing data, and decision support systems informed by captured weather data for four years now. We have learned a lot and will continue to do so. Of course, like any new technology or approach, we have discovered some challenges along the way. Weather conditions can often vary in just mere miles which necessitates a lot of hardware to capture those conditions. Publicly accessible stations exist but are often too far away to rely on accurate data. Interpolated models meant to guess weather conditions between publicly available stations have been very inaccurate, again, due to the distance between hardware. Today, deploying multiple cellular-enabled weather stations on an average-sized farm has remained cost prohibitive to many. Issues with station performance, maintenance, and sensor calibration can be all too common as well.
Not done yet. User interfaces for viewing and interacting with data can be cumbersome and frustrating. Even the ability of exporting data for use in Microsoft Excel is not necessarily a given. In situations where an Agronomist or farmer is working with multiple brands of hardware, data is likely fragmented across different digital tools and awkward to compile and compare. Actionable insights beyond immediate real-time readings are often untapped and can be abstract. In other words, while using the wind sensor to assess spray risk is certainly an immediate and worthwhile application, I'm also interested in, say, the relationships between nighttime temperatures and heat blasting in canola; hard to analyze without reliable, local temperature data at scale. We are seeing lots of work being done with turning basic sensor data into a generated recommendation for spray timing and disease risk. If accurate, models like this will add to the value of capturing weather data.
Raising a crop is a scientific endeavour. One of the foundations of scientific methodology relates to the practice of identifying, isolating, and controlling for variables. Rainfall, air temperature, and humidity are examples of variables and they are often missing pieces of the agronomy puzzle. Motivated by the potential insights into grain production and being cognizant of the challenges outlined above, G-Mac's AgTeam is piloting a strategic weather sensor infrastructure project unlike anything attempted previously. The project will be an evaluation of benefit to an installed grid of weather stations that aims to strike a balance between resolution and investment of hardware. Next spring, G-Mac's AgTeam will select three branches to install 50-60 weather stations to capture the conditions for those geographies. Customers will have the option to access the weather data under a subscription model. This approach will increase the accessibility for many more farmers interested in a local weather network without the upfront cost in hardware and responsibility of installation and maintenance. In addition to the weather data, G-Mac's AgTeam has partnered with a company whose expertise in predictive analytics has enabled them to build models for crop staging and assessing disease risk for sclerotinia and fusarium head blight. This information will be available remotely in a mobile app.
This is a new and exciting project being led by our Ag Innovation team. Fortunately, G-Mac's AgTeam has the resources, scale, and foresight to explore innovative initiatives such as this. As agriculture continues down a path of realizing previously unavailable insights derived from data analytics at scale, the capturing and digitizing of major agronomic data point fundamental to grain production and will only become more necessary (weather conditions, soil attributes, input management, yield, etc). There is no example of an agricultural weather sensor network at this scale and resolution in Western Canada. If this project shows promise in terms of value, interest, and execution, G-Mac's AgTeam will be further exploring onboarding more locations.