A precise assessment of porosity and permeability is crucial in characterizing petroleum reservoirs, as it is essential for predicting future behavior and guiding production strategies. This study utilizes the program developed by the Microfluidics and Flow in Porous Media Laboratory (LMMP) at PUC-Rio to evaluate how impactful noisy data or the lack of data can be in characterizing a petroleum reservoir. This program starts by simulating petroleum reservoirs based on desired porosity and permeability data. It then introduces noise or removes part of the data, proceeding to solve the inverse problem using the ES-MDA method to assess the impacts of noisy data and the absence of data in reservoir characterization. Three distinct scenarios were explored: a baseline case, used as a reference for comparison with others; a case involving the introduction of noise into the data; and a case implying data removal. For each scenario, two reservoirs were simulated, one with conventional properties and another with characteristics close to those of pre-salt reservoirs. At the end of the study, it was found that data exclusion did not influence problem resolution, whereas the introduction of noise resulted in a proportional increase in error.