Pasture Vibe is a computer model that calculates pasture growth rate. Supplied with climate data supplied nightly by NIWA (National Institute of Water and Atmosphere) Pasture Vibe can forecast up to 3 months of future daily pasture growth.

The model is very easy to set up. You only need to apply three inputs: your location, the type of farm (dairy or sheep and beef), and the sites fertility (high medium or low).

You will ask how can this be possible? Pasture growth is a complex response to the quality of soil and its fertility, the state of the pasture (the effect of existing pasture biomass), the level of soil moisture and the climate. We can access most of this information with no effort required from the farmer.

At the heart of Pasture Vibe is the soil water balance. Pasture Vibe assesses your water balance from the daily climate data received from NIWA and the amount of water the model calculates is used up by evapotranspiration from the pasture.

You can improve forecasts by adding any measurements of growth rate estimates you make; for example, if you use Farmax, all historical Farmax pasture growth calculations will be used by a “learning algorithm” to improve predictions. The learning algorithm adjusts certain parameters within the model and tests whether these lead to greater accuracy. It may be that your soil holds less moisture than the default selected for your site (during setup). In this case the algorithm observes that when soil water holding capacity is adjusted downward, the model predictions align better with your historical measurements.

With the default setting you would chose during setup, we have determined the model has a moderately high accuracy (measured by the percentage of pasture growth rate the model can effectively explain) of R²=0.65, or in other words the model explains 65% of all variation in pasture growth from one period to the next. Where you have added historical growth rate data the learning algorithm uses this to improve accuracy to R²=0.75, a value considered high by statisticians.