More learning. In this post I show how Tmax behaves, and how it can be an indicator of change but isn’t necessarily a predictor.
Results for January:
The rain we’ve had:
Here’s the January – March plot of the 3 day Tmin , Tmax and 3d rain anomaly indices, with the 3 day mean sea level pressure (scaled and inverted). I have made a small change- 3 day means are now centred. Predicted dates to watch are in green.
I have also graphed the actual one day data (without smoothing) for the month of January with the last 3 days of December and the first 3 days of February, which shows daily changes much more clearly. The 8 January change is visible. Changes can show up in Tmax as a drop in temperature, but not always. I’ve also indicated when the predictions were made.
Note the changes in temperature and pressure especially. A running temperature plot would be ideal for monitoring weather. Unfortunately I can only find 72 hours of data for BOM stations. Also note the small events around 18-19 and 21, and then the enormous pressure change and rainfall surge visible in the previous graphs, as the tropical depression from ex-TC Oswald moved through the region.
The rain started in Rocky on the night of 23 January, and continued further south through to 28 January. What was my prediction again? Oh yes, 23-27 ( on 21 January so not exactly a difficult prediction); 25 +/- 5 days (27 November- hmmm!); 22 +/- 2 days (20 September- check it! As was 14 January!).
The next change arrived on 31 January exactly on time, (20 September- “Late January- February (in the region 28 January-onwards)”) – and the instability persisted until storms cleared on 1 February. This was not the start of the wet season however. (The “big one” was the 23-27 event, not 30-31 as predicted on 27 November.)
For the year so far, 7 changes predicted, 7 out of 7 identified – and 3 of these were predicted 4 months previously.
It has been suggested that I tighten up my predictions as the uncertainty I used to give (+/- 3 days) meant nearly every day of a month was covered- I could never be wrong. In reply:
-I no longer quote uncertainties of +/- so many days (see 21 January post) but give a time period in which a weather event will occur.
-If you look at the predictions I made, you will see that all predicted weather events occurred well within 2 or 3 days of the date I specified, and most of them exactly right.
-I am trying to impress on my readers that weather is not random or chaotic: we just don’t understand it well enough yet. Rain occurs because it is part of a weather system, not through chance. Weather events are discrete and usually can be easily identified so can be matched with predictions. There will not be weather changes on every day of the month- they are usually 6-7 days apart but can be as close as 2-3 days apart.
I can predict weather events for sub-tropical Queensland with a fair degree of accuracy more than four months ahead, and can increase the accuracy about five weeks ahead.
I am now confident that my method gives an accurate forecast: however the dates remain approximate as we are talking about a weather influence making 4 complete orbits of the globe, and systems frequently slow down or stall. In certain seasons and different ENSO conditions there is more variation.
Therefore, I will give specific dates (in bold) but within a short time period. Some are certain to fall outside these limits.
Updated predictions for Sub-Tropical Queensland for February to July:
An enhancement (pressure change, temperature change, and/or rain) between these dates:
I’m having trouble interpreting data for March and April. Unstable conditions I think with frequent small changes?
Feb 28-1-2-3-4 (unsettled from 24 February to 9 March)
6-7-8-9 (any rain may start on 5 or 6)
28-29-30-31- 1 April.
4-5-6-7-8-9-10-11-12-13-14-15 (what will this mean?)
17-18-19-20 (Unsettled for the first 3 weeks)
7-8-9-10-11-12-13-14 (unsettled for the first 2 weeks)
June: (more instability)
SOI 30 day mean (to 2 February) is -2.8 (neutral).
Weekly NINO 3.4 Index (to 27 January) was -0.26 (neutral).
The Indian Ocean Dipole (to 27 January) was -0.09 (neutral).