Articles
FORECASTING THE AUSTRALIAN MACADAMIA CROP VIA MECHANISTIC AND STATISTICAL CLIMATE MODELS
Article number
773_23
Pages
165 – 172
Language
English
Abstract
Two levels of crop predictions are produced for the Australian macadamia industry.
The first is an overall long-term forecast, and is based on tree census data of growers in the Australian Macadamia Society (AMS). This data set currently covers around 70% of total production and is supplemented by our best estimates of non-AMS orchards.
A statistical model was developed to calculate average yields per tree using growers historical yields.
This model allowed for the effects of tree age, variety, year, region, and tree spacing, and explained 65% of the total variation in the yield per tree data.
The second forecast is an annual climate adjustment of the long-term estimates, taking into account the expected effects on production of the previous years climate.
This adjustment is based on historical yields, measured as the percentage deviance between expected and actual production.
The dominant climatic variables are observed temperature, rainfall, evaporation, solar radiation, and modelled water stress.
Initially, a number of models showed good agreement within the historical data, with jackknife cross-validation R2 values of up to 97%. However, forecasts varied quite widely between these alternate models.
Exploratory multivariate analyses and nearest-neighbour methods were used to investigate these differences.
For the first three years, the overall forecasts were in the right direction (when compared with the long-term expected values), but were not sufficiently accurate.
Recent results and plans for the future, including monitoring sentinel trees, will be discussed.
The first is an overall long-term forecast, and is based on tree census data of growers in the Australian Macadamia Society (AMS). This data set currently covers around 70% of total production and is supplemented by our best estimates of non-AMS orchards.
A statistical model was developed to calculate average yields per tree using growers historical yields.
This model allowed for the effects of tree age, variety, year, region, and tree spacing, and explained 65% of the total variation in the yield per tree data.
The second forecast is an annual climate adjustment of the long-term estimates, taking into account the expected effects on production of the previous years climate.
This adjustment is based on historical yields, measured as the percentage deviance between expected and actual production.
The dominant climatic variables are observed temperature, rainfall, evaporation, solar radiation, and modelled water stress.
Initially, a number of models showed good agreement within the historical data, with jackknife cross-validation R2 values of up to 97%. However, forecasts varied quite widely between these alternate models.
Exploratory multivariate analyses and nearest-neighbour methods were used to investigate these differences.
For the first three years, the overall forecasts were in the right direction (when compared with the long-term expected values), but were not sufficiently accurate.
Recent results and plans for the future, including monitoring sentinel trees, will be discussed.
Authors
R.A. Stephenson, D.G. Mayer
Keywords
macadamia, crop forecast, multiple regression, principal components analysis
Online Articles (45)