Although greenhouse agriculture can generate high crop yields, they vary due to spatiotemporal differences in incident light and photosynthesis. To elucidate these dynamics, multipoint analysis of hemispheric images and a photosynthesis model were used to visualize the spatiotemporal distribution of photosynthetic photon flux density (PPFD) and leaf photosynthetic rate (A) and compared these with strawberry fruit yield in a greenhouse. This method enabled successful estimation of spatiotemporal variability in PPFD and A with relative root mean square errors of 4.4% and 11.0%, respectively. PPFD, captured at ca. 2 m resolution, varied diurnally and seasonally based on sun position and external light intensity. A showed less spatial variability, because it is reduced by physical and physiological mechanisms in the leaves at excessive leaf temperatures and becomes saturated at high PPFD. Yield spatial variability was better explained by A than by PPFD. The association between A and yield weakened over the cultivation period (R² declined from 46% in winter to 12% in spring), thus suggesting that, over the cultivation period, factors such as photoassimilate availability replaced A as the primary limiting factor. The proposed method can be directly applied to other types of greenhouses, and the findings may facilitate spatiotemporal optimization in crop production, improving precision greenhouse agriculture.

虽然温室农业可以产生高作物产量，但由于入射光和光合作用的时空差异，它们会有所不同。为了阐明这些动态，使用半球图像的多点分析和光合作用模型来可视化光合光子通量密度 (PPFD) 和叶片光合速率 ( A ) 的时空分布，并将其与温室中的草莓果实产量进行比较。该方法能够成功估计 PPFD 和A的时空变异性，相对均方根误差分别为 4.4% 和 11.0%。PPFD，捕获于 ca。2 m 分辨率，根据太阳位置和外部光照强度在昼夜和季节变化。一个显示出较少的空间变异性，因为它在叶片温度过高时通过叶片中的物理和生理机制减少，并在高 PPFD 时变得饱和。A比 PPFD更好地解释了产量空间变异性。A与产量之间的关联在栽培期间减弱（R ²从冬季的 46% 下降到春季的 12%），因此表明，在栽培期间，光同化物有效性等因素取代了 A作为主要限制因素。该方法可以直接应用于其他类型的温室，研究结果可能有助于作物生产的时空优化，提高精准温室农业。