Instead of working with abstract dimension symbols, we can plug in the actual SI units, kilograms, metres and seconds, for each quantity and require the two sides of the equation to carry the same units.
Working purely in kilograms, metres and seconds, the only way every kilogram cancels and the metres and seconds reduce to \( \text{m}\cdot\text{s}^{-1} \) is with \( a = \dfrac{1}{2} \) and \( b = -\dfrac{1}{2} \), the square-root-of-tension over square-root-of-mass-per-length combination described by the first option.
So the correct answer is \( a = \frac{1}{2}, b = -\frac{1}{2} \), the first option.