Old Akro:
From memory, there was also a missed approach at YMIA that was not passed onto the crew.
Yes, there was a missed approach at Mildura, by a Qantas Link Dash 8. The Virgin aircraft was already overhead Mildura at the time and Qantas wasn't far behind. In fact the Qantas aircraft started its approach about a minute later. The report states that the Virgin crew asked the Dash 8 for an appreciation of the weather and was told they 'couldn't see anything'.
I also have a recollection that buried somewhere in the report is something to the effect that the BOM forecast the fog on the basis of satellite imagery rather than ground observation and that there was a ground observation that conflicted with one of the revised TAF's at the time of its issue.
The report states the following:
For regional airports, such as Mildura, the BoM relied on satellite imagery, in addition to other data, to provide indications of fog. However, it can be difficult to differentiate between fog and low cloud when using satellite imagery. The BoM reported that their review of the forecasting used for Mildura on the day of the occurrence showed that the conditions (including wind direction) were not conducive to fog developing at that time of day and were more consistent with the possibility of low cloud. As such, the TAF that was valid at the time reflected a temporary period of low cloud that was expected to last between 30 and 60 minutes (TEMPO), rather than fog.
The 'other data' refers to observations recorded by the equipment and the observer at the airport. The revised TAFs that were issued were only valid from 0000Z, ie after both aircraft had arrived overhead Mildura. The TAF that was valid at the time of their arrival had a TEMPO for low cloud at 600ft, but no fog.
The 2230Z and 2300Z METARs both showed good conditions, with the lowest cloud at 3400ft. Shortly afterwards, the instruments at Mildura began to indicate patches of low cloud at around 400ft and a satellite image indicated a bank of low cloud south of Mildura. The report states that the BoM believed the forecast for TEMPO deteriorations was still appropriate, based on those observations.
About 10 minutes later, the weather began to deteriorate rapidly, starting with low cloud at about 200ft AGL, followed by reduced visibility and then fog. The first report of the deteriorating conditions came from an aircraft that departed Mildura at 2316Z. The Qantas and Virgin aircraft were on a different frequency at the time and the report was not subsequently passed to them. A SPECI showing the low cloud was issued at 2318Z, followed by three more SPECIs in rapid succession at about 2330Z. During that time period, the visibility dropped from greater than 10km to 2100m. According to the report, the observer "advised the forecasting office that the mist and subsequent fog arrived rapidly from the south".
I'm no meteorologist, but it seems to me that the weather conditions were quite good until shortly after 2300Z, with no indications of what was to come. The weather then deteriorated rapidly after about 2315Z. By that time, both the Virgin and Qantas aircraft had already diverted from Adelaide, on the assumption that the conditions at Mildura were still as reported at 2300Z. The BoM started issuing SPECIs as soon as the conditions began to deteriorate, but they were not relayed to either aircraft and neither was the airborne report from the aircraft that departed Mildura at 2316Z.
To be honest, I can't really pin the 'blame' on the BoM apart from the inaccuracy of the initial TAFs for Adelaide and Mildura, which did not predict fog. That said, in my experience, predictions of fog are rarely accurate at some airports, which is probably an indication of how difficult it is to forecast such conditions, rather than ineptness on the part of the BoM. I guess they're damned if they do and damned if they don't, in that they'll be pilloried for getting it wrong either way. In this case, it appears that everyone was caught out by the rapidly changing conditions.
In my view, the most easily solved causal factor in this incident is the lack of reports passed to the aircraft by ATC once the conditions started to deteriorate. If those reports had been relayed immediately, the crews might well have made a decision to return to Adelaide and attempt an autoland, which would obviously have been the lesser of the two evils. In that light, I believe the major problem lay with Airservices procedures, as highlighted by the ATSB's report.
I totally agree that a time line in the report would help readers to put a few things in perspective!