For y>0y>0, the Poisson kernel of the upper half-plane is

Py(t)=1πyt2+y2,t∈R.P_y(t)=\frac1\pi\frac{y}{t^2+y^2},\qquad t\in\mathbb R.

It is positive, integrates to 11, and is as a function of t+iy∈Ht+iy\in\mathbb H.

Approximate identity

As y↓0y\downarrow0, the family PyP_y concentrates at the origin. Therefore Py∗f→fP_y*f\to f in standard LpL^p regimes and at Lebesgue points. The convolution Py∗fP_y*f is the of ff.

Scaling

Py(t)=y−1P1(t/y)P_y(t)=y^{-1}P_1(t/y). This scaling explains the equivalent formula ∫f(x+ty)/(1+t2) dt/π\int f(x+ty)/(1+t^2)\,dt/\pi used in linewise extensions.

References
  1. Elias M. Stein and Rami Shakarchi, Complex Analysis, Princeton University Press, 2003. DOI record.