The 23 GHz emission lines from the NH _ { 3 } rotation inversion transitions are widely used to investigate the kinematics and physical conditions in dense molecular clouds .
The line profile is composed of hyperfine components which can be used to calculate the opacity of the gas ( 16 ) .
For intrinsic linewidths of a few km s ^ { -1 } , the 18 magnetic hyperfine components blend together to form a line profile composed of five quadrupole hyperfine lines .
If the intrinsic linewidth exceeds one half of the separation of these quadrupole hyperfine components ( \sim 5 - 10 km s ^ { -1 } ) these five lines blend together and the observed linewidths greatly overestimate the intrinsic linewidths .
If uncorrected , these artificially broad linewidths will lead to artificially high opacities .
We have observed this effect in our NH _ { 3 } data from the central 10 pc of the Galaxy where uncorrected NH _ { 3 } ( 1,1 ) linewidths of \sim 30 km s ^ { -1 } exaggerate the intrinsic linewidths by more than a factor of two ( 8 ) .
Models of the effect of blending on the line profile enable us to solve for the intrinsic linewidth and opacity of NH _ { 3 } using the observed linewidth and intensity of two NH _ { 3 } rotation inversion transitions .
By using the observed linewidth instead of the entire line profile , our method may also be used to correct linewidths in historical data where detailed information on the shape of the line profile is no longer available .
We present the result of the application of this method to our Galactic Center data .
We successfully recover the intrinsic linewidth ( \langle \Deltav _ { int } \rangle \approx 15 km s ^ { -1 } ) and opacity of the gas .
Clouds close to the nucleus in projected distance as well as those that are being impacted by Sgr A East show the highest intrinsic linewidths .
The cores of the “ southern streamer ” ( 13 ; 3 ; 4 ) and the “ 50 km s ^ { -1 } ” giant molecular cloud ( GMC , Güsten , Walmsley , & Pauls ( 10 ) ) have the highest opacities .