Reading the Guardian article and presenting the facts...
Abares has built a model that spits out the impact from market fluctuations and climate variations
This paper presents farmpredict a data driven micro-simulation model of Australian broadacre farms based on the Australian Agricultural and Grazing Industry Survey (AAGIS). Farm production and financial data from AAGIS are combined with site-specific climate data including various measures of rainfall, temperature and soil moisture. A statistical model is estimated linking the production of outputs (e.g., wheat, beef cattle, wool etc.), the use of inputs (e.g., fuel, fertiliser, labour etc.) and changes in farm stocks (e.g., livestock and grain) with farm fixed inputs, input and output prices, climate variables and other control variables. The model is estimated using a non-parametric machine learning method, which combines a gradient boosted regression tree algorithm (xgboost) with multi-target stacking (two-stage regression). The resulting model can be used to forecast or simulate production, financial outcomes and stock changes for individual farms given scenarios for climate conditions and commodity prices. The performance of the model is evaluated via cross-validation. Simulation results are presented showing the types of farm responses to climate and price shocks produced by the model.
So what has been developed by ABARES is a tool that if you specify a climate and a commodity value it will forecast the impact on a farm based on those inputs. Link to the full document https://www.agriculture.gov.au/sites...farms%20v2.pdf
A paper has also been written that uses this model to generate an output of farm profitability since 2000
https://www.agriculture.gov.au/sites...rms_v1.0.0.pdf
It does tend to state the bleeding obvious, less rain means less income but it is worth a read as the paper attempts to quantify how agricultural enterprses can best manage risk. There is a section that the Guardian has pounced on that compares farm outputs 1950 - 1999 and 2000 - 2018 but taken out of the context of the paper's intent i.e. to illustrate how changing climate and commodity prices impact farm profitability these figures mean little.
The Guardian article finishes with a tilt at government policy and references "typical farming enterprises and farming household support" in its final paragraph as a subtle pull at the reader's heartstrings to conjure up the image of starving farmers. I would contend that the structural changes to agriculture since the 1950s has had a far greater effect on agriculture than climate change. We own what would have been a typical farming enterprise in the 1950s that supported multiple generations of the same family since the 1880s. That farm is now nothing more than a hobby farm, there is no way that it would sustain my family, these changes have nothing to do with climate change but are reflective of the massive amount of change that has taken place in the industry and the improved standard of living in that time.
There is also bias in comparing the era between the 1960s and the late 1970s when Australia really did ride on the sheeps back and owning a farm was a licence to print money. Those days ended sometime in the 1970s unfortunately and agriculture is no different to any other industry.
By the way Mick the report does not state that climate change is reducing incomes 22% year on year, it states that the simulation shows some broadacre enterprises have a 22% reduction in average income between 1950 - 1999 and 2000 - 2019 average income. There is possibly some bias in that statistic as well, can you reliably compare rainfall over a 50 year period with rainfall over a 20 year period?
Regards,
Tote


 
						
					 
					
					 Originally Posted by DiscoMick
 Originally Posted by DiscoMick
					
 
				
				
				
					 Reply With Quote
  Reply With Quote 
						
					 
				 
					
					 
						
					 
						
					 
						
					 
						
					 
			
Bookmarks