Objective  Photochemical reflectance index (PRI) plays an important role in accurately estimating photosynthetic light use efficiency (LUE). However, at different levels, such as leaf, canopy and landscape levels, the relationship between PRI and LUE and their influencing factors are different. The spectra obtained by the sensor are pixel and canopy spectra. The PRI-LUE relationship model of the leaf level cannot be directly used for canopy-level data. Therefore, the canopy level PRI needs to be scale-converted to obtain the leaf level PRI.

  Method  In this paper, the leaf-level PROSPECT model was used to simulate the reflectance and transmittance of the leaves considering different biochemical parameters, and then the leaf-level PRI and the simple ratio PRI (denoted as SR-PRI) were calculated. Secondly, the obtained leaf-level reflectance and transmittance were input as parameters to 4-scale model to get canopy reflectance under different LAIs, and the canopy-level PRI and SR-PRI were calculated. Finally, the regression analysis of PRI and SR-PRI in canopy and leaf level was carried out under different LAIs. The sensitivity of the influencing factors on PRI and SR-PRI at different scales was analyzed.

  Result  PRI and SR-PRI of the canopy level have a good linear relationship with that of the leaf level. SR-PRI is generally better than PRI at the relationship of both levels, and the coefficients R² is positively correlated with LAI.

  Conclusion  The 4-scale model is feasible for scale conversion between canopy and leaf for PRI and SR-PRI. Canopy level PRI and SR-PRI can be transformed to leaf level by the functions established at different LAIs efficiently.