Spatio-Temporal Modelling Informing Wolbachia Replacement Releases in a Low Rainfall Climate

Releases of Aedes aegypti carrying Wolbachia bacteria are known to suppress arbovirus transmission and reduce the incidence of vector-borne diseases. In planning for Wolbachia releases in the arid environment of Jeddah, Saudi Arabia, we collected entomological data with ovitraps across a 7-month period in four locations. Herein, we show that mosquito presence in basements does not differ from that of non-basement areas of buildings. In modelling mosquito presence across the study sites, we found the spatial structure to be statistically significant in one of the four sites, while a significant spatial structure was found for egg production data across three of the four sites. The length scales of the spatial covariance functions fitted to the egg production data ranged from 143 m to 574 m, indicating that high productivity regions can be extensive in size. Rank-correlation analyses indicated that mosquito presence tended to persist from the dry to wet season, but that egg production ranks at locations could reverse. The data suggest that, in Jeddah, the quality of the local environment for breeding can vary over time. The data support the feasibility of dry season releases but with release numbers needing to be flexible depending on local rates of invasion.

Forecasting Multiple Groundwater Time Series with Local and Global Deep Learning Networks

Time series data from environmental monitoring stations are often analysed with machine learning methods on an individual basis, however recent advances in the machine learning field point to the advantages of incorporating multiple related time series from the same monitoring network within a 'global' model. This approach provides the opportunity for larger training data sets, allows information to be shared across the network, leading to greater generalisability, and can overcome issues encountered in the individual time series, such as small datasets or missing data. We present a case study involving the analysis of 165 time series from groundwater monitoring wells in the Namoi region of Australia. Analyses of the multiple time series using a variety of different aggregations are compared and contrasted (with single time series, subsets, and all of the time series together), using variations of the multilayer perceptron (MLP), self-organizing map (SOM), long short-term memory (LSTM), and a recently developed LSTM extension (DeepAR) that incorporates autoregressive terms and handles multiple time series. The benefits, in terms of prediction performance, of these various approaches are investigated, and challenges such as differing measurement frequencies and variations in temporal patterns between the time series are discussed. We conclude with some discussion regarding recommendations and opportunities associated with using networks of environmental data to help inform future resource-related decision making.