Policies to regulate severe surface ozone pollution in cities in India are challenging to develop, due to the complex dependence on precursor emissions of volatile organic compounds (VOCs) and nitrogen oxides (NOx), non-linear chemistry leading to ozone formation, and very limited spatial and temporal surface air quality monitoring. Ratios of space-based observations of formaldehyde (HCHO), an intermediate oxidation product of VOCs, and nitrogen dioxide (NO2) have been used to characterize the sensitivity of surface ozone production to precursor emissions of VOCs and NOx, but interpretation of these depends on the local oxidation regime. Here we develop an improved approach in which we discretize the data into background HCHO due to methane and other long-lived VOCs (regression intercept) and the local relationship (regression slope) between HCHO associated with reactive VOCs and NO2. We apply this to TROPOMI HCHO and NO2 tropospheric columns oversampled to higher spatial resolution than the native pixel resolution of the instrument over the ten most populous cities in India. We use GEOS-Chem to characterize the ozone production regimes and then apply this updated interpretation of the relationship between HCHO and NO2 to the oversampled TROPOMI columns to identify the most effective strategies for regulating ozone and whether these should vary seasonally and spatially.