Fire has shaped the environment and has been important for human cultural development. In this paper, we propose to study past fire events using ecological modelling. For instance, the ecology of fire can help us to understand and interpret archaeological problems related to past settlement patterns or environmental scenarios. Variation in fire regimens are related to both, vegetation and precipitation fluctuations. Recently, we have model past ecosystem dynamics using remote sensing in the Ambato Valley (Catamarca NW Argentina) ranging from 442 to 1998 CE. Our aim here is to use remotely sensed vegetation data to enhance our understanding of environmental disturbance in the period 2000-2011. We characterised the spatial-time dynamics of the annual NDVI as an indicator of vegetation functioning. Then we related the NDVI dynamics to precipitation and fire events in an arid highland environment in the area. Further, we analysed the vegetation data (monthly NDVI, MODIS/TERRA satellite, 1km2 pixels), and the climate data: annual precipitation. Then we calculated the NDVI annual average of every pixel and the NDVI anomalies of every year over the studied period. Lastly, we related NDVI data with annual precipitation and compared the NDVIs prior to and after known fire events in this period. On a spatial scale, the results show that the NDVI values were (a) low in shrublands and in cultivated areas, (b) medium in grasslands and piedmont forest with anthropic impact, and (c) high in highland forests. Within the studied time-period, extreme positive and negative anomalies were detected. The precipitation inter-annual variations were greater than the NDVI inter-annual variations, thus demonstrating that in some areas of the valley the horizontal precipitation can make important contributions to the ecosystem humidity. Extreme negative anomalies were observed the year of fire and fire scars at least for the next two years. These results demonstrate the relation between structure and function of vegetation, precipitation and fire. Understanding these relations can enable us to explain results when hindcasting ('predicting' what happened during past episodes of climate change) palaeoenvironmental conditions and fire events, thus helping us to interpret different archaeological contexts related to fire events.