Please use this identifier to cite or link to this item: https://hdl.handle.net/2440/113427
Full metadata record
DC FieldValueLanguage
dc.contributor.advisorOstendorf, Bertram Franz-
dc.contributor.authorBracho Mujica, Gennady-
dc.date.issued2018-
dc.identifier.urihttp://hdl.handle.net/2440/113427-
dc.description.abstractAgricultural management needs relevant climate information to reduce the climate uncertainty and support crucial management decisions. Risk profiles of modelled crop yields (cumulative probability curves) are effective tools for summarising long-term yield variability, exploring the benefit and limitations of agricultural management decisions and serve to quantify the impact of future climate conditions. However, modelling reliable crop yield and risk profiles requires continuous, accurate, and long-term (>100 years) local weather records for rainfall, temperature, and solar radiation, which are not always available. This study aimed to systematically assess spatial and temporal factors that limit the accuracy of risk profile of modelled crop yields. The specific objectives were (1) to analyse if and to what degree short time series of weather data can be used to provide reliable risk profiles, (2) to test how simple adjustments of high-quality local data can be used to extrapolate risk profiles across broad climatic regions, and (3) to address a combination of sparse spatial coverage of climate data and short daily weather observations. Here we focused on the Australian grain-belt selected on the basis of the availability to high-quality, long-term climate data, widely used and calibrated process-based crop model (APSIM, Agricultural Production Systems sIMulator). To examine the sensitivity of risk profiles of modelled crop yields to the temporal coverage of the climate data, 15 wheat-growing sites were selected based on their proximity to weather stations with high-quality daily weather records for the last 100 years (baseline period). Risk profiles were constructed using variable temporal coverages and compared with risk profiles obtained for the baseline period. Results indicated a decline of modelled wheat grain yields, particularly for the last three decades. They also highlight the interactions between model complexity and data demand. The sensitivity of the risk profiles to record length was increased in models accounting for severe frost and heat events. The second research objective of this study addresses spatial extrapolation and explores to what extent a simple method for adjusting daily weather data using seasonal and monthly factors could produce robust estimates of risk profiles at a continental scale. Adjustment factors were calculated as the difference in long-term average of a given climate variable between 49 test sites and the reference site. Risk profiles modelled with observed weather data were compared with those modelled with adjusted data. Simple adjustments of both precipitation and temperatures produced reliable risk profiles in 80% of the sites. This study implies that for regions with limited availability of high-quality climate data, simple scaling of climate inputs can provide basic climate data for modelling and generating robust spatial patterns of risk profiles of crop yield. The third objective addresses the realistic scenario of using modern, process-based crop models, which are data hungry, in data sparse environments. Models that can capture combinations of potential climate and management impacts on food production require complex climate data that are either not available or difficult to access at high spatial detail and/or temporal extent for many parts of the world. Here, we assess the sensitivity of the risk profile accuracy to the temporal coverage of the climate data combined with spatial adjustments of daily weather data for risk profile modelling purposes. In this case, adjustment factors were determined using a variable temporal coverage at every study site. Risk profiles were modelled using observed and adjusted weather data covering different periods. Results indicated that although adjustment factors are very sensitive to the record length of the climate data, it was possible to produced reliable risk profiles with only 10-30 years of climate data. This research has increased our understanding of the sensitivity of risk profiles to the temporal and spatial aspects of climate data availability. It highlights the usefulness of risk profiles to characterise spatial and temporal patterns of yield and will help to improve agricultural management under climate uncertainty.en
dc.subjectResearch by publicationen
dc.subjectclimate risken
dc.subjectclimate data qualityen
dc.subjectcrop modellingen
dc.subjectrisk profileen
dc.subjectAPSIMen
dc.titleModelling crop yields and climate risk under limited climate data conditionsen
dc.typeThesesen
dc.contributor.schoolSchool of Biological Sciencesen
dc.provenanceThis electronic version is made publicly available by the University of Adelaide in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. This thesis may incorporate third party material which has been used by the author pursuant to Fair Dealing exceptions. If you are the owner of any included third party copyright material you wish to be removed from this electronic version, please complete the take down form located at: http://www.adelaide.edu.au/legalsen
dc.description.dissertationThesis (Ph.D.) (Research by Publication) -- University of Adelaide, School of Biological Sciences, 2018en
Appears in Collections:Research Theses

Files in This Item:
File Description SizeFormat 
Bracho Mujica2018_PhD.pdf7.76 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.