#Background
##General description
Under the REDD + program (Reduce Emissions from Deforestation and forest Degradation), Mexico is committed to the development of a robust nationwide system for monitoring activity data. For this purpose the use of the Inventatio Nacional Forestal y de Suelos (INFyS) and satellite remote sensing products is proposed. The correct classification of land cover and land cover changes are an input of vital importance in the study of the activity data. Given the country's size, you need an automated method of processing that amount of information.
In this context MAD-Mex system arises (Monitoring Activity Data for the Mexican REDD+ program).The purpose is to provide the ability to process large amounts of data in a reasonable time.
##Scopes
Being an automated system, MAD-Mex products are conditioned to availability and quality of data input. The weather conditions determine the behavior of classification algorithms. The presence of clouds, fog and snow reduces the amount of useful information that you can extract from the images, and in certain areas of the country is almost impossible to get clean images.
Also, algorithms require quality training data. Today, the volume of training data with which account is suboptimal. Generating training data for a country with the size and complexity of Mexico requires a significant amount of resources.
On the other hand, statistical techniques offer countless of alternatives for the study and production of products. The MAD-Mex system explores just a few methods. The study of other alternatives allow improved product quality, however, this is again conditioned to restrictions of time and resources.
##Clarifications
The MAD-Mex products are preliminary, the system is constantly improving and comments and suggestions are welcome.