Infrastructure Engineering - Theses
Permanent URI for this collection
Search Results
1 - 1 of 1
-
ItemTowards improved irrigation scheduling through sensitivity analysis and remote sensing of crop coefficients( 2022)Agriculture faces significant challenges to increase food and fibre production under increasingly variable climate and uncertain supply of resources such as water. At present only 37% of the fresh water delivered to agricultural crops is used by crop in part due to inefficient management of irrigation. Therefore, optimal irrigation scheduling is important for saving water while maintaining crop productivity. To improve irrigation scheduling, it is important to identify the key input factors influencing the efficacy of the scheduling. In this study, the FAO56-based soil water balance irrigation scheduling method has been selected as one of the most widely-applied methods (Allen et al., 1998). Firstly, realistic ranges of the uncertainties of ten selected input factors of the method, including weather, soil, crop, and management factors, have been evaluated to assess their impact on irrigation scheduling. Then, the sensitivity of the irrigation scheduling to the uncertainty of each input factor is calculated using the Sobol’ global sensitivity analysis method. Results show that FAO56 crop coefficient, which has an expected average uncertainty of 20%, has the highest impact on irrigation scheduling. The sensitivity analysis was performed for eight different climates in Australia, which showed that the ranking of the influential factors on irrigation scheduling did not change with climate conditions, making crop coefficient the most sensitive factor in all climates. Since crop coefficient is an important factor in irrigation scheduling, FAO56 and remote-sensing-based Irrisat crop coefficients were compared with measured crop coefficients in one maize field in Australia and two lucerne fields in New Zealand. The Irrisat-derived dynamic crop coefficient values reproduced temporal changes in crop coefficient better than the FAO56 values. FAO56 crop coefficients were within a reasonable range for most periods of crop growth but they could not capture their temporal changes. However, Irrisat underestimated crop coefficients during the mid-season and end-season for maize, likely due to the saturation effect of the Normalised Difference Vegetation Index (NDVI) used to derive the dynamic crop coefficient. This problem was less significant in the lucerne fields since they feature lower biomass ranges. Overall, while FAO56 coefficients captured reasonable values in magnitude, the Irrisat-derived crop coefficient is superior in detecting temporal changes when it is available. Hence, using both of them in consideration of their advantages and disadvantages can help reduce the uncertainty in crop coefficients and ultimately save water in irrigation scheduling. Future research should focus on developing effective assimilation approaches to combine these different sources of crop coefficients, and thus ultimately improve the accuracy while reduce uncertainty of their estimates.