You can navigate to other sections of the report by folllowing these links:

  1. What Overseer is
  2. The Overseer farm file
  3. What Overseer produces
  4. How the Overseer engine works
  5. Uncertainty in model results
  6. Overseer Glossary



The Overseer science model

Overseer is a science model made up of three components: 

  • the user input fields where a farm is described,
  • the engine that models the nutrient movements for that farm through a series of calculations, and
  • the reports which display the analysis for interpretation (and are stored in the database).
All three components work together to undertake scenario analyses of different nutrient outcomes for a farm to inform management decisions.

The engine is made up of a series of sub-models that calculate the flow of seven major farm nutrients – Nitrogen (N), Phosphorus (P), Potassium (K), Sulphur (S), Calcium (Ca), Magnesium (Mg), and Sodium (Na) and greenhouse gases.

 

What Overseer covers

The Overseer model has been expanded over time (see development history timeline) to cover a wide range of different farm enterprises and management practices in New Zealand. It can model mixed and single enterprise farms of varying intensity and with or without irrigation.

Farm types currently modelled by Overseer include:  

Pastoral enterprises including dairy, beef, sheep, dairy goats and deer. These include farms using fodder crops (fodder beets, kale, rape, swedes, turnips), forage crops (ryegrass, barley, oats, maize, rye corn, triticale) and animal housing.

Permanent fruit crop enterprises including avocado, kiwifruit, apples, peaches and grapes.

Horticulture enterprises including:

  • Green vegetables – broccoli, brussel sprouts, cabbage, cauliflower, lettuce, spinach
  • Legume vegetables – beans, lentils, peas
  • Root vegetables – kumara, potatoes, carrots, beets, parsnips
  • Other vegetables – onions, sweetcorn, squash tomatoes.

Arable crop enterprises including:

  • Grain crops – barley, maize, oats, wheat
  • Seed crops – ryegrass, clover.

Overseer science model calibration

To ensure Overseer is generating expected results, like other models it is calibrated against measured data. In 2012, when the nitrogen (N) pastoral grazing model embedded in Overseer was revised to capture newly developed science on urine patch dynamics, it was calibrated against farmlet-scale N leaching measurements.
Since then there have been a number of potentially significant changes to the model, including the incorporation of the new irrigation sub-model (version 6.2) and changes to feed allocation for seed crops and ME modelling.
The new FM software includes facilities to test detailed measured data against modelled results (including individual sub-models within Overseer) making it possible to assess the modelling against more data (including lysimeter trials) that have become available since 2012.
Overseer Limited currently has a N-loss through leaching calibration project underway and one in preparation.

1.     Calibration of the N-model for pastoral systems

The pastoral N-model calibration (undertaken by AgResearch) includes assessment of the drainage, background and urine-patch submodels as well as whole model N-loss to the root zone results. This will establish a close relationship between measured data and model results.
There are approximately 20 farmlet data sets for pastoral systems (Dairy and Sheep and Beef) covering Southland (4), south Otago, central Otago, Otago, Canterbury, Manawatu (3), Waikato (7) Rotorua and The Bay of Plenty. There are approximately 30 Lysimeter trial datasets – these include Templeton, Horotiu, Lismore, Taupo pumice, Matuara, Harihari and Tokanui soils. A comprehensive list will be reported later in the year and the aim is to publish the calibration exercise in a peer reviewed journal. Any changes to the model to ensure the model is delivering expected results are due to be identified by the end of June 2019 and implemented soon after.

2.     Calibration of the N-model for arable systems

We are currently preparing to undertake a calibration of the N-model for Arable systems using the data that has become available from a fluxmeter network that has been in place in New Zealand for the last few years. This data is provided by the Ministry for the Environment, Foundation for Arable Research, Plant and Food Research and other funding partners.
This includes 12 farms in Canterbury, Manawatu, Hawkes Bay, Waikato and Pukekohe regions with arable and vegetable rotations including grains and seeds, onions, maize, potatoes, beetroot and leafy green vegetables – the results still need to be assessed before we can confirm they are able to be used successfully in this process. We are also working with Zespri to assess data from Kiwifruit trials.
Subject to confirming arrangement with required experts we are planning to undertake this work in the next 6 months

 

Overseer science model version numbers

As with any software, Overseer is updated at regular intervals - to fix known problems, improve or add features and add new science to the science model. When updates are made to the science model, the model version number changes. The version number can be found on the top right of the farm analysis page in OverseerFM. 
The current version number is 6.3.1. The number format represents the scale of change

X
is the major version number. It only increases when there’s a major change to the design or structure of the model. The last update was in 2012, when Version 6 was released to include the urine patch model.
Y
is the minor version number. It increases when functionality changes are made, for example, new science or sub-models are added. The last minor version change occurred in 2018, when changes were made in how feed was allocated from grazed crops.
Z
is the patch version number, and increases when minor changes are made, for example bug fixes.


 

Using Overseer results

Overseer adds value to farm management decisions because it predicts the outcomes of changes in farm management on nutrient loss. Because OVERSEER provides an individual analysis of a farm, it helps users identify the best ways to manage that farm and mitigate potential losses of nutrients into the environment.

The following summary identifies the farm management areas that have the most influence on nutrient loss estimates:
 

Category Variable Effect on calculated Ability of farmers to influence
Nitrogen Phosphorus
Animal Stocking rate × × ×
  Species ×   ×
  Gender ×   ×
  Production and Reproduction policy ×   ×
  Sheep/Beef ratio ×   ×
  Management × × ×
Fertiliser Fert N rate ×   ×
  Fert N form ×   ×
  Fert P rate   × ×
  Fert P form   × ×
  Timing × × ×
Effluent management Amount × × ×
  Timing × × ×
General Rainfall × ×  
  Irrigation rate × × ×
  Topography   ×  
Pasture Clover level ×   ×
  Pasture type ×   ×
  Pasture ME ×   ×
Soil Natural drainage × ×  
  Artificial drainage × × ×
  Anion storage   ×  
  Soil properties × ×  
  Olsen P   × ×
Supplements Farm grown supplements ×   ×
  Imported supplements ×   ×
Crops Type × × ×
  Management × × ×


Go to next section:  The Overseer farm file