Turbulence plays an important role in the structure and dynamics of the galaxy and influences various processes therein including star formation. In this work, we investigate the large scale turbulence properties of the external spiral galaxy NGC 5236. We combine the VLA multi-configuration archival data with the new GMRT observation to estimate the column density and line of sight velocity fluctuation power spectra for this galaxy over almost two decades of length scales. The energy input scale to the ISM turbulence is found to be around $6$ kpc. Power spectra of the two-dimensional turbulence in the galaxy's disk follow a power law with a slope $-1.23\pm0.06$ for the column density and $-1.91\pm0.08$ for the line of sight velocity. The measured power spectra slopes strongly suggest in favour of a compressive forcing with a steady energy input of $\sim 7 \times 10^{-11}$ erg cm$^{-2}$ sec$^{-1}$. We conclude that much of these originate from the gravitational instabilities and self-gravity in the disk. This is the first and most comprehensive study of turbulence statistics for any external spiral galaxy.

Turbulent dynamics generate random fluctuations in density and velocity in the Interstellar Medium (ISM) of spiral and dwarf galaxies. Observationally, H~{\sc i} 21-cm radiation provides a good probe of these stochastic processes and helps us to know more about their nature and generating mechanism. Structure function, auto-correlation function, power spectrum are some of statistical estimators of these fluctuations which have been used in literature, whose studies helps in understanding the generating mechanisms of these structures. In this thesis we are interested in estimating the column density and velocity power spectrum at scales as large as 10 Kpc, to understand the energy associated with these structures and the type of forcing of the turbulence. Estimating power spectrum from observations can be done either by using the directly measured quantity visibility or by reconstructing the sky brightness distribution (or image) from the visibilities. By using simulated observations of a model galaxy with known power spectrum we find that the visibility based estimator reproduce the true power spectrum. The image based estimator has a scale dependent bias that is highly correlated to the incompleteness in the baseline coverage of the interferometer. Interestingly, we found that the noise in the reconstructed image has zero mean and hence the locally averaged quantities estimated from the reconstructed image are unbiased. We implement the visibility moment estimator as discussed in Dutta et al. (2015) and estimate the line of sight velocity fluctuation of the spiral galaxy NGC6946. Our result strongly indicate presence of large scale turbulence in the spiral galaxies. We conclude that we would need to have higher signal to noise H~{\sc i}observations to definitively comment on the generating mechanisms of the large scale turbulence.

Radio interferometric data are used to estimate the sky brightness distributions in radio frequencies. Here we focus on estimators of the large-scale structure and the power spectrum of the sky brightness distribution inferred from radio interferometric observations and assess their efficacy using simulated observations of the model sky. We find that while the large-scale distribution can be unbiasedly estimated from the reconstructed image from the interferometric data, estimates of the power spectrum of the intensity fluctuations calculated from the image are generally biased. The bias is more pronounced for diffuse emission. The visibility based power spectrum estimator, however, gives an unbiased estimate of the true power spectrum. We conclude that for an observation with diffuse emission the reconstructed image can be used to estimate the large-scale distribution of the intensity, while to estimate the power spectrum, visibility based methods should be preferred.