Current capacity of a microstrip, how do you compute?

anybody have a reliable formula for computing the current capacity of a microstrip line?

any conductor "current capacity" is defined by its cross-section, resistivity and allowed temperature increase.
Microstrip lines are often made by etching, so their real cross-section is unknown, and temperature increase is dependent on substrate gradient and conductivity, etc.

Instead of current capacity it is a rule of thumb to allow a maximum RF power, usually ~10 W CW max. Make sure that your stripline connects a well-matched load- standing waves generate local overheating that can destroy the stripline even under a lower current than your estimate gives.

Most small RF microstrip uses 1/2oz copper, meaning that weight of copper spread out over 1 sq ft. area. The 1 oz. option and thicker is always available, and more. You can have it very thick if you like.

The microstrip impedance model has the thickness as a factor, but it is a very small effect unless you are using edge-coupled features, coplanar WG or other bits that involve close approaches.
For me, as example a 0.508mm (20mil) board has 17.5micron (0.68mil) thick copper at the start.
After the vias are plated 25um, the side-effect is the microstrip tracks are thicker 42.5um (1.67mil).

Be aware that if the thing you are making has resonant parts, they can run a huge circulating current.

Consider adhesion. If you are getting into combiners and splitters and power amplifiers using low-loss substrates, they may have a temperature let-go specification.
Over time, have come across several on-line aids to help this issue. Some are better than others. Many are devoted to FR4 and the switch-mode power supply industry.

Here is a typical example --> LINK **broken link removed**
A Google-style search will yield more..