This work provides a spatio-temporal statistical modeling for egg and adult Aedes mosquitoes count data with consideration of environmental data in the context of mosquito epidemiology. For a given spatio-temporal incomplete mosquito count data, we derive predictions assuming that all of the spatio-temporal dependence can be accounted by potential factors influencing the development of Aedes mosquitoes such as, rainfall, temperature (including many delay) and waste data. In this paper, after a data analysis with the entomological and environmental data, we apply a LASSO regression to perform a variable selection strategy (we are in the presence of a large number of explanatory variables). We highlight the relevant factors that explain the abundance of our egg and adult mosquitoes count data. We define a spatio-temporal risk ratio which is a probability of exceeding a given threshold value of mosquito abundance. We propose two spatio-temporal modeling approaches for the Aedes mosquitoes' count data. The first is based on a spatio-temporal kernel smoother and the second on a generalized additive model. The paper conclude with a detailed discussion that follows not only, the spatio-temporal prediction and model performance measures, but also the obtained spatio-temporal risk ratio in order to highlight potential space-time areas of threshold exceedances of mosquitoes' abundance where health services could apply vector surveillance and control measures. We also discuss a forthcoming work concerning the theoretical developments of a spatio-temporal model for simulation purposes and an R Shiny application called May'Aedes, that aims to be a flexible and efficient tool to predict spatio-temporal risk ratio.