Compared with larger nations, small island developing states (SIDS) are disproportionately affected by natural disasters relative to size and frequency. Social, environmental and economic complexities are integral in making SIDS more vulnerable to such catastrophes. Floods and droughts have drastic impacts on societies; however, few studies have been performed in the Pacific region to reduce risks. Hydrological indices are useful to understand vulnerabilities and can be used to improve knowledge of spatial and temporal distribution of drought and floods, thereby improving preparedness. For this research, the following hydrological indices were tested for the Fiji region: Standardised Precipitation Index (SPI), Standardised Precipitation Evapotranspiration Index (SPEI), Effective Drought Index (EDI) and self-calibrated Palmer Drought Severity Index (PDSI). Because of the importance of agriculture in Fiji, the performance of each index was verified using two plant productivity indices: Enhanced Vegetation Index (EVI) and Normalised Difference Vegetation Index (NDVI). It was concluded that the EDI is the best-performing index and that precipitation is the principle variable in monitoring hydrological extremes in island topographies. Further, a meta-analysis was carried out, and it was concluded that wavelet-transformed artificial neural networks (WANNs) have the best performance to forecast drought indices. Therefore, we employed an artificial neural network (ANN) as well as a WANN model to forecast the EDI at 1 month lead time using climate indices. As water is critical in the germination phase of sugarcane, it is essential that new crops receive water within a week of being planted. Interviews with farmers revealed that a short term forecast (1-3 month) in advance will assist them to make informed agronomic decisions. The models showed promising results in predicting EDI; however, both the models showed average performance in predicting extreme events. Instances of over-prediction and under-prediction were noted for both the models from the categorical verification. We also used multivariate statistical techniques to carry out spatial drought modelling. The results were not better than those of the neural networks; however, multivariate techniques have added advantages such as identifying relationships between response and explanatory variables, which can be used with other techniques. In conclusion, neural networks can be used to implement an operational hydrological extreme monitoring system in Fiji. Further testing and optimisation to better predict the extreme events will be useful for informing the public; this area warrants future investigation