In this study, the application of artificial intelligence to monthly and seasonal rainfall forecasting in Queensland, Australia, was assessed by inputting recognized climate indices, monthly historical rainfall data, and atmospheric temperatures into a prototype stand-alone, dynamic, recurrent, time-delay, artificial neural network. Outputs, as monthly rainfall forecasts 3 months in advance for the period 1993 to 2009, were compared with observed rainfall data using time-series plots, root mean squared error (RMSE), and Pearson correlation coefficients. A comparison of RMSE values with forecasts generated by the Australian Bureau of Meteorology's Predictive Ocean Atmosphere Model for Australia (POAMA)-1.5 general circulation model (GCM) indicated that the prototype achieved a lower RMSE for 16 of the 17 sites compared. The application of artificial neural networks to rainfall forecasting was reviewed. The prototype design is considered preliminary, with potential for significant improvement such as inclusion of output from GCMs and experimentation with other input attributes.

No abstract

The catalytic cracking and skeletal isomerization of n‐hexenes on 60/80 mesh ZSM‐5 zeolite were studied in the temperature range 350–405°C. By applying the time‐on‐stream theory, the products of the reaction were identified as primary, secondary or both according to their optimum performance envelope (OPE) curves on product selectivity plots. The products of cracking were found to be almost exclusively mono‐olefins and those in the range C3‐C5 were found to be stable primary, or primary plus secondary products. No C1 was found, and only traces of C2 as ethylene. The observed product distributions can be explained by a dimerization‐cracking mechanism with no product species having more than twelve carbon atoms. The probability of a fragment undergoing further cracking before desorption increased with temperature and the observed initial selectivities must be corrected to account for this process. Methylpentenes, formed as unstable primary products, were the main isomers produced by skeletal rearrangement, with those derived from more stable carbenium ions predominating. Paraffins, coke and aromatics were found in small amounts only.

The catalytic cracking of n‐alkenes on ZSM‐5 zeolite at 405°C can occur both by a monomolecular mechanism and a bimolecular process. In the latter, cracking is preceeded by dimerization. We show that pentenes are cracked exclusively by the bimolecular process. The dominant cracking mechanism for n‐hexenes also requires initial dimerization, although a small proportion (<19%) of the total cracking may proceed by a monomolecular process. Cracking of n‐heptenes is predominantly monomolecular, with only 13% of the total occurring via an initial dimer formation. The cracking of n‐octenes and n‐nonenes can be interpreted by assuming a monomolecular mechanism only. Thus it appears that at 405°C olefins smaller than C6 are stable with respect to direct cracking and must dimerize before a species is formed which is unstable enough to crack. No molecular hydrogen was produced in any of the cracking reactions reported here in the range of conversions studied.

The decomposition of alkaline hydrogen peroxide solutions at 20°C has been studied in the presence of both supported iron catalysts and in systems with iron initially in solution. Studies with an iron-alumina supported catalyst showed the decomposition reaction was first order with respect to total peroxide concentration, while studies with alkaline Fe3' produced more complex behavior. This has been attributed to the presence of at least two distinct catalytically active iron species. The first species is highly active and gives rise to high initial rates of reaction. A decrease in concentration of this species is observed together with an increase in concentration of a second, less active, iron species. The catalytic behavior of this "aged iron species was found to be very similar to that of the supported iron catalyst.

Catalytic cracking of a product from Fischer-Tropsch synthesis, consisting of -paraffins (92.2%) and monomethyl paraffins (7.8%) In the range C5-C28, has been studied on HY and HZSM-5 at 405 °C. Cracking increased the ratio of isomeric monomethyl/linear paraffins in the range C20-C14 on HZSM-5, whereas the ratio decreased in products formed on HY. The spectrum of low molecular weight cracking products on HZSM-5 was concentrated in the range C3-C8, while on HY this range was broader (C3-C10). These product distributions are explained by assuming that the paraffin molecules penetrate zeolite pores in a lengthwise orientation until an active site is encountered. Cracking is initiated by formation of a pentacoordlnated carbonium ion on the third or subsequent carbon atom in the chain to produce a saturated fragment and a residual carbenium ion. Linear product paraffins are preferentially formed on

Brisbane, the capital of Queensland, Australia, has flooded periodically and catastrophically, most recently in January 2011. Official seasonal rainfall forecasts failed to predict the floods. Since winter 2013, the Australian Bureau of Meteorology uses a general circulation model, the Predictive Ocean Atmosphere Model for Australia (POAMA), to make official seasonal rainfall forecasts presented as the conditional probability of rainfall being greater or less than the long-term median rainfall. We show that a more skilful forecast can be made using an artificial neural network (ANN), a form of statistical modelling based on artificial intelligence. A Jordan recurrent neural network with one hidden layer was implemented, using genetic optimization of inputs. For the sites of Gatton and Harrisville, in the Brisbane catchment, monthly rainfall forecasts from the ANN show lower root mean square errors than forecasts from POAMA. These rainfall forecasts from the ANN model were further improved by using inputs of independently forecast values for climate indices including the Southern Oscillation Index, the Interdecadal Pacific Oscillation, Pacific sea surface temperature anomalies (Niño 3.4) and also atmospheric temperature. The results presented here represent a first attempt at independently forecasting climate indices using an ANN model for the Australian east coast.

No abstract