Osborne, John (2011) Seasonal climate forecasting for sugar cane: an economic investigation of harvest management. Masters (Research) thesis, James Cook University.

Preview

PDF (Thesis) - Submitted Version Download (6MB)

Abstract

Climatic variability represents a major cost to agriculture. Farmers worldwide are forced to make many management decisions on the basis of uncertain future climate. This 'moving target' of climate leads to losses as management decisions must be tailored to suit a wide spectrum of possible climatic outcomes. Seasonal climate forecasts can narrow down the range of expected outcomes, allowing farmers to focus and improve climate-related decisions.

While forecast technology has developed rapidly, it is only through corresponding advances in our ability to understand and to use these forecasts that these advances will become valuable (Meinke et al., 2006). This thesis builds upon existing research on the value of seasonal forecast information value by analysing the economic potential of new seasonal forecasts for sugarcane farmers. Cane, one of the world’s major tropical crops, has been little considered in this field of study.

The specific problem I address is the use of new long-lead seasonal climate forecasts to reschedule the annual harvesting operations. By forewarning the industry of potentially heavy spring rainfall, harvesting of crops can be shifted earlier, avoiding potential damage. The case study for this research is the Herbert River cane growing district which surrounds the town of Ingham, in North Queensland. Due to the volatility of climate and price, and the presence of precursors for non-rational farmer psychology, the region offers a particularly good locale in which to compare biophysical, economic and psychological influences on information value.

The research employs a bio-economic decision model, which was developed and tested in three stages. First, I constructed an expected profit decision model for the farmer’s annual choices regarding the harvest. Feeding into this model was data on farm outcomes derived from extensive crop simulations (using APSIM: an Agricultural Production Simulation Model, developed by the CSIRO in Australia), and a financial model of farmer profit (developed for this thesis). Second, the model was extended into an expected utility model that allowed for farmer risk aversion. This model was used to examine the sensitivity of information value to a range of other factors, including price and uncertain future climate.

Forecast values were modest, though comparable with other similar studies. Perfectly accurate forecasts gave only slightly higher information values than the existing forecast, indicating that prediction of ENSO phase is of secondary concern to the high degree of intra-phase volatility in the ENSO signal. Results indicate that information value varies with soil type and will thus be spatially disaggregated across the region and even within farms. The value of information was relatively invariant to the risk attitude of the farmer and to input and output price fluctuations. Both these results appear to be the natural consequence of the relatively flat profit function faced by the farmer in this decision problem. This highlights the importance of considering flat profit functions when analysing on-farm decision problems; a fact which has been generally ignored by the forecast valuation literature.

Finally, I undertook a comprehensive review of the different assumptions that researchers often make about biophysical, economic and psychological factors when assessing information values. Several of these assumptions were then varied in realistic ways – the primary aim being to determine which types of factors had the most/least significant impact on information values. The findings indicated that introducing even a relatively minor non-rational element into the farmer's decision behaviour in certain years produced larger deviations in information value than did alterations in assumptions about other biophysical or economic factors. In addition, when the farmer engages in non-rational behaviour information value is often negative.

The research is unique in at least four aspects: (1) it develops an agronomiceconomic model specific for sugar cane which is further integrated with a wet weather model to simulate resource-constraint costs (2) it provides the first decision model estimates of forecast value in sugar cane, (3) it simulates forward looking forecast use, thus identifying situations in which forecasts may have negative information values, and (4) it considers a realistic psychological bias induced by the forecast itself and compares the resulting changes in forecast value to changes induced by biophysical and economic factors.

Item ID:	29156
Item Type:	Thesis (Masters (Research))
Keywords:	Herbert River; North Queensland; economic value; seasonal climate forecasts; forecast value; sugar cane; decision modelling; risk attitude; information value
Related URLs:	http://eprints.jcu.edu.au/20783/ http://eprints.jcu.edu.au/19499/
Additional Information:	Publications arising from this thesis are available from the Related URLs field. The publications are: Chapter 4: Osborne, J.A., Stoeckl, N.E., Everingham, Y.L., Inman-Bamber, N.G., and Welters, R. (2011) The economic value of conditioning harvest start date on long-lead seasonal climate forecasts. Proceedings of the 2011 Conference of the Australian Society of Sugar Cane Technologists 2011 Conference of the Australian Society of Sugar Cane Technologists. , 4-6 May 2011, Mackay, QLD, Australia. Chapter 6: Everingham, Yvette L., Stoeckl, Natalie E., Cusack, Justin, and Osborne, John A. (2012) Quantifying the benefits of a long-lead ENSO prediction model to enhance harvest management: a case study for the Herbert sugarcane growing region, Australia. International Journal of Climatology, 32 (7). pp. 1069-1076.
Date Deposited:	05 Sep 2013 22:09
FoR Codes:	14 ECONOMICS > 1402 Applied Economics > 140201 Agricultural Economics @ 100%
SEO Codes:	97 EXPANDING KNOWLEDGE > 970114 Expanding Knowledge in Economics @ 50% 91 ECONOMIC FRAMEWORK > 9104 Management and Productivity > 910499 Management and Productivity not elsewhere classified @ 50%
Downloads:	Total: 645 Last 12 Months: 15
	More Statistics