New Computer Fund

Saturday, July 14, 2012

Analyse This

UPDATE:  UAH oceans included at the end of the post:

Anyone can look at data.  If a chart is going up, it is going up.  Does it have to keep on going up?

You can draw a line on the chart to show it is going up.  You can start at different time to show it is going up less or more from that point.  Once you get near the end of the chart there is no more data. You can't predict the future, just look at the past, just see what happened in the past.  You can extend that line or curve you fit to the data in the chart, and extend those into the future as a guess of what may be coming.  If you pick a starting point that shows a future more like you want, you have "cherry picked" your starting point.  That's a no no.



So say you want to not fool yourself or others and just want the facts.  In the chart above I picked all possible 60 month trends for the UAH lower troposphere global temperature data.  If I wanted to show the most positive trend for a 60 month period, I would pick the 60 months ending in about 1989.  The most negative 60 month trend would end in about 1994.  From those two points the trends decrease.  For the whole period, the trend of the trends is upward a little bit.



If you do the same analysis for the Northern Hemisphere, the Southern Hemisphere and the Tropics, you see there are differences and similarities.  They don't do the same things at the same times.  The most negative 60 month trend for the tropics ended in about 2001, for the SH, the most negative land trend ended in 1994 but the oceans and total had their most negative trends en in about 2001.  For the NH, the most negative ocean trend was in about 2009.

If you were investing all your money in UAH Lower Troposphere temperature anomalies, would you buy or sell.

Bonus question

What about Sea Surface Temperature Anomalies?

 This is the same type of linear regression analysis done with the UAH ocean data.  Each series is 60 month linear regression ending at the date for each month of the series.  The beginning is obviously limited to 59 month after the beginning of the data.  Note that as the individual portions of the oceans synchronize something would be happening to the energy transfer inside the oceans.  With adjacent oceans sections synchronized, there would be less internal energy transfer or a more stable rate of internal energy transfer.  The oceans would be more likely to cool if they are in equilibrium or steady state.  This method is just a quick way to determine an estimate of the degree of synchronization.  

No comments:

Post a Comment