We present new diagnostic tools for distinguishing supernova remnants (SNRs) from HII regions. Up to now, sources with flux ratio [S II]/Hα higher than 0.4 have been considered as SNRs. Here, we present the combinations of three or two line ratios as more effective tools for the separation of these two kinds of nebulae, depicting them as 3D surfaces or 2D lines. The diagnostics are based on photoionization and shock excitation models (MAPPINGS III) analysed with Support Vector Machine (SVM) models for classification. The line-ratio combination that gives the most efficient diagnostic is: [O I]/Hα -[O II]/Hβ -[O III]/Hβ. This method gives 98.95% completeness in the SNR selection and 1.20% contamination. We also define the [O I]/Hα SNR selection criterion and we measure its efficiency in comparison to other selection criteria.

We present a systematic study of the Supernova Remnant (SNR) populations in the nearby galaxies NGC 45, NGC 55, NGC 1313, and NGC 7793 based on deep H$\rm {\alpha }$ and [S ii] imaging. We find 42 candidate and 54 possible candidate SNRs based on the [S ii]/H$\rm {\alpha }$>0.4 criterion, 84 of which are new identifications. We derive the H$\rm {\alpha }$ and the joint [S ii]-H$\rm {\alpha }$ luminosity functions after accounting for incompleteness effects. We find that the H$\rm {\alpha }$ luminosity function of the overall sample is described with a skewed Gaussian with a mean equal to $\rm \log (L_{H\alpha }/10^{36}\, erg\, s^{-1})=0.07$ and $\rm \sigma (\log (L_{H\alpha }/10^{36}\, erg\, s^{-1}))=0.58$. The joint [S II]-H$\rm {\alpha }$ function is parameterized by a skewed Gaussian along the log([S ii]$\rm /10^{36}\, erg\, s^{-1}) = 0.88 \times \log (L_{H\alpha }/10^{36}\, erg\, s^{-1}) - 0.06$ line and a truncated Gaussian with $\rm \mu (\log (L_{[S\, II]}/10^{36})) = 0.024$ and $\rm \sigma (\log (L_{[S\, II]}/10^{36})) = 0.14$, on its vertical direction. We also define the excitation function as the number density of SNRs as a function of their [S ii]/H$\rm {\alpha }$ ratios. This function is represented by a truncated Gaussian with a mean at -0.014. We find a sub-linear [S ii]-H$\rm {\alpha }$ relation indicating lower excitation for the more luminous objects.

We analyzed the massive star population of the Virgo Cluster galaxy NGC 4535 using archival Hubble Space Telescope Wide Field Planetary Camera 2 images in filters F555W and F814W, equivalent to Johnson V and Kron-Cousins I. We performed high precision point spread function fitting photometry of 24353 sources including 3762 candidate blue supergiants, 841 candidate yellow supergiants, and 370 candidate red supergiants. We estimated the ratio of blue to red supergiants as a decreasing function of galactocentric radius. Using Modules for Experiments in Stellar Astrophysics (MESA) isochrones at solar metallicity, we defined the luminosity function and estimated the star formation history of the galaxy over the last 60 Myr. We conducted a variability search in the V and I filters using three variability indexes: the median absolute deviation, the interquartile range, and the inverse von-Neumann ratio. This analysis yielded 120 new variable candidates with absolute magnitudes ranging from MV = −4 to −11 mag. We used the MESA evolutionary tracks at solar metallicity to classify the variables based on their absolute magnitude and their position on the color-magnitude diagram. Among the new candidate variable sources are eight candidate variable red supergiants, three candidate variable yellow supergiants and one candidate luminous blue variable, which we suggest for follow-up observations.