That PX constraint comes from MEGAN 3.2, but the MEGAN implementation in CMAQ should be able to use NOAH LSM if you comment out these lines in BDSNP_MOD.F
C ! make sure it only runs with PX version
IF( .NOT. PX_LSM ) THEN
MESG = “BDSNP Soil NO is only compatible with PX version”
CALL M3EXIT( PNAME, JDATE, JTIME, MESG, 2)
END IF
This check was left in place pending further testing. The WSAT values are different for NOAH. There is also some inconsistency in the design. PX uses soil moisture for the top 1 cm, while in the NOAH top layer is 10 cm thick. The amount of NO produced by the BDSNP model is dependent on the frequency of drying/wetting in that top layer. The original implementation of BDSNP used 2 cm soil depth from GEOS-Chem, so possibly less pulsing than what you would see with PX but more pulsing than you would get with NOAH.
Proceed with caution and make sure the soil NO emissions look reasonable for your application.