A series of scaling studies attempting to correlate the H(high)-mode power threshold (P-TH) with global parameters have been conducted. Data from these discharges is also being used to look for dependence of P-TH on local edge parameters and to test theories of the transition. Boronization and better operational techniques have resulted in lower-power thresholds and weaker density scaling. Neon impurity injection experiments show that radiation also plays a role in determining P-TH. A low-density threshold for the L(low)-H(high) transition has been linked with the locked mode low-density limit, and can be reduced with the use of an error field correcting coil. Highly developed edge diagnostics, with Spatial resolution as low as 5 mm, are used to evaluate how the power threshold depends on local edge conditions. Preliminary analysis of local edge conditions for parameter scans of n(e), B-T, and I-p in single-null discharges, and the X-point imbalance in double-null discharges show that, just before the transition to the H mode, the edge temperatures near the separatrix are approximately constant at 100<T-i<220 eV and 35<T-e<130 eV, even though the threshold power varied from 1.5 to 14 MW. During a density scan, the edge ion collisionality, nu(*i), varied from 2 to 17, demonstrating that a transition condition as simple as nu(*i)=const is inconsistent with the data. The local edge parameters of n(e), T-e, and T-i do not always follow the same global scaling as P-TH. Therefore, theories of the L-H transition need not be constrained by these scalings. (C) 1996 American Institute of Physics.
