Thermals are generated by buoyant air. Air becomes buoyant because it is less dense than the surrounding air. Differential heating of the ground causes differential heating of the layer of air above the ground. The warmer air expands becoming less dense and thus buoyant. Perhaps surprisingly the addition of water vapour also makes air buoyant. This is because water vapour is 5/8th as dense as air. So in general we look for areas that will be hotter and or have moisture added as likely sources of thermals. Beware that to much moisture can have a detrimental effect, whereas a little can have a very beneficial effect. A good way to get a feel for this is just take a walk and observe the temperature - if the air feels hotter then the surface you are walking over is a likely THERMAL GENERATOR. Classic generators include dark ground, burnt areas, tarmac roads/carparks, etc. Sand reflects heat so is bad. Some crop paddocks get surprisingly hot, whereas others are cool. The vital point to remember is the concept of differential heating - what you want is a contrast. By a contrast I mean an area that will get hot next to or better still surrounded by area which are cool. The edges of forests, river banks and lake edges are all potentially good. Areas which heat up fast are good at the beginning of the day. Areas which heat up more slowly can be good towards the end of the day - for example you often find thermals over forests later in the day.



    Just because air is buoyant does not guarantee a thermal. Just as water can cling to a ceiling until a drip forms so buoyant air can cling stubbornly to the ground. Before anything happens it must be TRIGGERED to release. A good analogy is to imagine that the ground is the ceiling of a steam room. Anywhere you would expect water to drip from so you can expect thermals to trigger from. In practical terms look for high points. The flatter the ground the less significant the high point. In the mountains ridge tops are good but in the flatlands treelines, houses, rockpiles and even lonely telegraph poles all act as triggers.

    Wind complicates the picture. Buoyant air can drift with the wind along the ground until it is triggered far from where it was generated. In this case sloping thermals result and you will tend to fall out the downwind side unless you continually centre the core by flying upwind due to the fact that the thermal rises ~200 fpm faster than you do because even though you are climbing up from the ground you are always sinking down through the air. Alternatively the internal turbulence of the moving mass of buoyant air may cause it to trigger independent of ground features - in this case the thermals will be, perhaps surprisingly, vertical because the source is moving with the wind.

    Wind also influences the nature of thermals. Strong winds encourage thermal triggering resulting in short lived bubble type thermals.

    The air in areas which are protected from the wind can continue to get hotter for longer before triggering - this often results in strong lee side thermals. Crop paddocks often hold onto their heating airmass for longer and can be better thermal generators than the classic ploughed paddock in windy conditions.

    Experience shows us that whenever the wind blows thermals will generally be far longer downwind than they are wide, often with several cores lined up downwind.

    On any given day thermals tend to remain similar in nature, unless of course there is a large change in conditions.

Bank Angle
    Good thermal pilots do not necessarily bank more or less than average pilots. What they do do is bank as much as is required to position their gliders in the core of the thermal.

    Although some authors labour on about optimum bank angles the rule is simple. Bank up enough to stay in the core! Experiment. More bank => better climb? => continue banking it up. If more bank leads to a slower climb then make shallower turns.

    We expect small bubbles near the ground so expect to have to bank it up. Late in the day wider thermals are the norm so shallower turns are usually the go.

    So how do you centre the core? There are several methods, of which I will mention two.

    The standard method is to tighten your turns when the lift drops off (to bring the glider quickly back into the best lift) and to flatten the turns as the lift increases (to fly into the best lift). The pro method is to fly into the thermal, feel the glider react to the air and then crank (bank it up) when you hit the core - more on this next.

Feel
    Some pilots have a better natural feel than others, but don't despair its really quite straight forward.

    As you correctly point out variometers have some lag. Sure some are faster and more sensitive than others but as a tool FOR CENTERING thermals they basically suck. Heresy to some no doubt but still true.

    In a big gaggle at any given comp you will see pilots circling around many different points. Why is it so? They can't all be in the core. The fact that some pilots climb much faster proves the point. These eccentric circles result from what I believe is a total over dependence on variometers combined with the standard method for coring a thermal described above.

    OK, here is what happens. Consider a glider flying in a straight line at 24mph (36km per hour / 10 metres per second) straight across the centre of a thermal. It will take this glider 9 seconds to traverse a thermal 90m in diameter. Lets say this thermal has a 30m or 3 second wide core in the centre. The glider enters thermal and is accelerated upwards. After a lag of say 2 seconds the glider ascends far enough for the variometer to note a change in air pressure and indicate a climb. Military studies indicate it will take about 1 second for the pilot to process this information by which time the glider has entered the core. A further 2+1 seconds elapse while the glider accelerates/pressure changes/pilot assimilates change. Just as the pilot notes he is in the core he in reality actually flies out of it. Using classical theory he decides to bank it up when the vario indicates a drop off in lift. This occurs 2+1 seconds later just as the glider exits the thermal. The pilot now banks up the glider which takes a further 2 seconds due to glider response lag. At this stage the pilot is actually 20m past the entire thermal! You can continue this description on indefinitely however the point is this:

    The classical method of centering a thermal will only work if there is no lag in variometer response, pilot (processor) response, and glider response.

    So now we come to the secrets of thermaling - visualisation and feel.