1. ECMWF proposal to introduce new parameters for clarification.
https://groups.google.com/a/wmo.int/forum/#!topic/cbs-ipet-drmm/mjqgaFE3gSE
That proposes new parameters:
* 0-1-93 Relative humidity with respect to water
* 0-1-94 Relative humidity with respect to ice
There is already "0-1-1 Relative humidity" but ECMWF propses adding
*both* water and ice because 0-1-1 was used for RH computed using the
"virtual" saturation (not commonly-understood term) defined by water-ice
mixture.
2. JMA uses such RH, at least in the global model. Saturated vapor
pressure is linear combination of water and ice, and the mixing ratio
changes linearly from -15 C to 0 C.
Sources:
* Para 3.2.5 of NWP Outline 2013
http://www.jma.go.jp/jma/jma-eng/jma-center/nwp/outline2013-nwp/index.htm
* Para 3.5.2(4) of NWP text #43 (in Japanese)
http://www.jma.go.jp/jma/kishou/books/nwptext/43/chapter3.pdf#page=28
3. Large-scale condensation scheme
The reason to have such virtual saturation should be large-scale
condensation.
For example following article describes a formulation that leads to JMA
schme.
W.K. Tao, J. Simpson, and M. McCumber, 1989: An Ice-Water Saturation
Adjustment. MWR 117 231-235.
http://journals.ametsoc.org/doi/abs/10.1175/1520-0493%281989%29117%3C0231:AIWSA%3E2.0.CO%3B2
The minimum temperature of presence of water droplets was undefined in
Tao et al (1989). Smith (1990), cited by JMA Outline, reports -15 C was
a choice by tuning.
R.N.B. Smith, 1990: A scheme for predicting layer clouds and their water
content in a general circulation model. QJRMS 116 435-460.
http://onlinelibrary.wiley.com/doi/10.1002/qj.49711649210/abstract
