Mosquitoes are dangerous vector organisms that spread diseases to millions of people worldwide, causing nearly one million deaths each year. This study serves to identify larvae of the malaria-spreading Anopheles genus of mosquitoes in North America while fueling methods to refine or “clean” the NASA GLOBE Observer data set and promoting Citizen Science. The GLOBE Observer app has facilitated mosquito research, allowing Citizen Scientists to report mosquito breeding grounds and the presence of larvae. In conjunction with other data, such as landcover or water, mosquito activity can be tracked and their effects can be mitigated. Citizen Science is often considered highly inaccurate, with the reasoning that almost anyone, trained expert or not, can contribute to data collection with varying levels of precision. To improve the accuracy and credibility of such Citizen Science data sets–in this case, the GLOBE Observer database–a sample of 155 unique mosquito observations were pulled from the database. Using this sample, trained classifiers and mosquito experts reclassified each reported observation to gauge Citizen Scientists’ accuracy in identifying Anopheles larvae. Using this reclassified data set, a convolutional neural network (CNN) was created as a machine learning (ML) solution to automatically identify a given larva photo as Anopheles or non-Anopheles. This model contains roughly 20% of the larval images in the GLOBE database, which were deemed usable for the training model. Keywords: Anopheles, Citizen Science, convolutional neural network, image classification, mosquito habitat.