State diagram (i.e. 12 micro-regions) of ripe banana was mapped by measuring and modeling its freezing point, glass transition, maximal-freeze-concentration conditions, solids-melting, and BET-monolayer. At 20 degrees C, the BET-monolayer moisture was observed as 0.044 g/g dry-solids, and decreased with the increase of temperature. Un-freezable water was found as 0.26 g/g sample and the maximal-freeze concentration temperature (Tm') was observed as -34.5 degrees C. The freezing point and solids-melting peak were modeled by Chen's and Flory-Huggins models, respectively. Browning of banana stored at different moisture and temperature (i.e. at different micro-regions) were measured as a function of storage time and modeled with first order reaction kinetics. The variation of reaction rate constant was analyzed based on the glass transition, water activity and macro-micro region concepts. At a specific moisture content, reaction rate constant showed a shift (i.e. sample containing freezable water) or change in slope (i.e. sample containing un-freezable water), when plotted as a function of temperature. However, it was difficult to find any validity above or below glass transition (or BET-monolayer) when all data points (i.e. all moisture and temperature) were plotted (i.e. rate constant with moisture or temperature). Arrhenius plot at moisture content 0.04 g/g sample showed two linear regions (i.e. below and above critical temperature 45 degrees C) with activation energy values of 105.3 and 25.1 kJ/mol, respectively. Universal validation was difficult to achieve, thus the rate constants within different micro-regions were empirically correlated with moisture content, storage temperature, BET-monolayer and glass transition temperature (p < 0.001). (C) 2017 Elsevier Ltd. All rights reserved.