Minh Trinh

PoliSci PhD @MITPoliSci
Asst. Prof @purduepolsci

Authoritarian Politics, Accountability, Pol. Behavior of Development

[Minh Trinh @mdtrinh_]

@GarethNellis yes absolutely!

[Minh Trinh @mdtrinh_]

@GarethNellis generally though, calibration that works at the pixel level ends up underperforming when aggregating to calculate total/mean nightlight for whole countries/regions, probably b/c errors for pixels close to each other are not random and accumulate into big biases

[Minh Trinh @mdtrinh_]

@GarethNellis most recently the EOG released a new series called DMSP-like VNL https://t.co/AUzpE2wjYg. This is the most straightforward data to use. There is also DMSP data extended to 2012-2018 straight from EOG (but still substantively different from older DMSP data)

[Minh Trinh @mdtrinh_]

@GarethNellis Hi @GarethNellis I worked on harmonizing DMSP/VIIRS and am happy to help!