Hematite (alpha-Fe2O3) is widely recognized as a promising candidate for the production of solar fuels via water splitting, but its intrinsic optoelectronic properties have limited its performance to date. In particular, the large electrochemical overpotential required to drive the water oxidation is known as a major drawback. This overpotential (0.4 - 0.6 V anodic of the flat band potential) has been attributed to poor oxygen evolution reaction (OER) catalysis and to charge trapping in surface states but is still not fully understood. In the present study, we quantitatively investigate the photocurrent and photovoltage transient behavior of alpha-Fe2O3 photoanodes prepared by atmospheric pressure chemical vapor deposition, under light bias, in a standard electrolyte, and one containing a sacrificial agent. The accumulation of positive charges occurring in water at low bias potential is found to be maximum when the photocurrent onsets. The transient photocurrent behavior of a standard photoanode is compared to photoanodes modified by either a catalytic or surface passivating overlayer. Surface modification shows a reduction and a cathodic shift of the charge accumulation, following the observed change in photocurrent onset. By applying an electrochemical model, the values of the space charge width (5-10 nm) and of the hole diffusion length (0.5-1.5 nm) are extracted from photocurrent transients' amplitudes with the sacrificial agent. Characterization of the photovoltage transients also suggests the presence of surface states causing Fermi level pinning at small applied potential. The transient photovoltage and the use of both overlayers on the same electrode enable differentiation of the two overlayers' effects and a simplified model is proposed to explain the roles of each overlayer and their synergetic effects. This investigation demonstrates a new method to characterize water splitting photoelectrodes-especially the charge accumulation occurring at the semiconductor/electrolyte interface during operation. It finally confirms the requirements of nanostructuring and surface control with catalytic and trap passivation layers to improve iron oxide's performance for water photolysis.