Tariq, I differ from Simon's opinion in a few points...
1 It's better to use 1/12dB smoothing to get a better appreciation from the graphs as then we know what we're actually looking at. Var Smoothing is variable, so we can't accurately judge the true height of the peaks and the Q factor of the resonances.
2 The waterfall trace hasn't been generated for a long enough time for your quite resonant room (I suggest 800ms or more will give a better picture); the plot needs to be scaled in time to show the resonances decaying almost to completion (at least -40dB of decay should be shown). The fact that the entire band 37Hz to 65Hz and the peaks at 28Hz and 74Hz decay so slowly indicates the requirement for bass traps in the room.
3 The resonance is at 37Hz (not 39Hz) corresponds to a room dimension of 4.6m. However it's indicated as only +8dB wrt the octave surrounding it, and it's actually entirely in line with the amplitude 100Hz - 200Hz, so, in the greater scheme of things it isn't really at all prominent. In fact the prominence of the 100Hz - 200Hz band looks to me to be more of an issue, causing an overblown 'plumminess' to the bass; rather than the specific 37Hz local peak.
4 The minor 74Hz resonance (corresponding to 2.3m) is most likely to be the primary mode of the floor to ceiling distance, rather than a harmonic of the 37Hz mode.
5 There is another notable resonance at 28Hz, corresponding to a room dimension of 6.1m, but at this frequency the anechoic bass response of the speakers is beginning to fall off quite a bit. Extending the water fall to 800ms (in fact, even 1s or more may be needed to show the decay) will show how important it is to deal with this resonance.