To determine the mechanisms of global warming, better understand the feedback process between the atmosphere and the terrestrial ecosystem, and more fully understand the role of terrestrial ecosystem carbon cycle transitions, we must make a quantitative spatial assessment of the various, highly mutable factors that control terrestrial carbon fluxes. We quantitatively analyzed spatial and temporal net ecosystem production (NEP) factors from 2001 to 2010 at a spatial grid resolution of 1 km in Japan. After conducting six sensitivity study runs for time-variable inputs, we calculated NEP anomaly and correlation coefficients between the default and each sensitivity study run to quantitatively examine the contribution of target inputs on NEP. The primary factor controlling the annual NEP was input related to radiation (solar radiation and albedo), and the area averaged significant correlation coefficient was 0.38, dominating 91.8% of the area. The correlation coefficient in most areas was significantly high, but in mountainous areas (>1000 m), it was low. Input related to temperature (air temperature and land surface temperature) was the second most influential factor (R2 = 0.10, area = 6.8%), and water had a significant correlation in just 28.2% of all pixels. We determined that the primary spatial and temporal NEP factors in Japan were radiation and temperature, with the dominant position gradually rotating by season among three factors. We reaffirmed the continuing importance of high spatial resolution observations to qualitatively monitor the spatial and temporal changes of the various control factors.