Progress in our understanding of the solar corona properties is highly dependant of the emipirical or semi-empirical determination of the plasma fundamental parameters, such as magnetic field, density and temperature. However, there is no direct measurements of such quantities; the integration along the line of sight considerably complicates the interpretations of the observations, due to the superimposition of structures with different properties. To avoid this ambiguity, there exist several tools, including the Differential Emission Measure (DEM) and the tomography reconstruction technique. The former provides the quantity of emitting material as a function of the temperature, whereas the latter is able to reconstruct the three dimensional distribution of the coronal emissivity. Coupling these two techniques leads to a three dimensional diagnostic of the temperature and density. The inversion code used in this work is currently one of the two codes in the world able to perform this coupling. The method described in this work has been developed in order to estimate the robustness of the spectroscopic diagnostics using the DEM formalism, using a new characterisation method taken into account the different uncertainty sources involved in the inversion process. Using a probabilistic approach, this technique is able to calibrate a priori the DEM inversion problem and thus allows to study the inversion behavior and limitations in the context of simple DEMs models. The advantage of this method is its ability to provide confidence level on the reconstructed DEMs computed from real data. First applied to the SDO/AIA (Atmospheric Imaging Assembly) imager in the case of simple models able to represent a variety of plasma conditions, our results show that DEM inversion of isothermal or near-isothermal plasmas is robust, whereas the multithermal solutions are less accurate but also biased to secondary solutions. We also applied the method to the Hinode/EIS (EUV Imaging Spectrometer) spectrometer, using a power law DEM, typical of active regions DEM, from which the slope provides important constraints related to the coronal heating frequency. Our results point out that the different uncertainty sources are currently too high to allow exploitable measurements of this frequency. The last part is dedicated to the three-dimensional reconstructions obtained by coupling tomography and DEM tools, focusing on polar structures. First reconstructions obtained using AIA data, our results allow to study the evolution of the temperature and density as a function of altitude, showing polar plumes denser and hotter than their surrondings.
