Using Field Spectroradiometer to Estimate the Leaf N/P Ratio of Mixed Forest in a Karst Area of Southern China: A Combined Model to Overcome Overfitting

The ratio between nitrogen and phosphorus (N/P) in plant leaves has been widely used to assess the availability of nutrients. However, it is challenging rapidly accurately estimate leaf N/P ratio, especially for mixed forest. In this study, we collected 301 samples from nine typical karst areas Guangxi Province during growing season 2018 2020. We then utilized five models (partial least squares regression (PLSR), backpropagation neural network (BPNN), general (GRNN), PLSR+BPNN, PLSR+GRNN) plants based on these samples. also applied fractional differentiation extract additional information original spectra each sample. results showed that average was 17.97. Plant growth primarily limited by areas. sensitive had wavelengths ranging 400–730 nm. prediction capabilities can be ranked descending order as PLSR+GRNN, PLSR, GRNN, BPNN when considering both accuracy robustness. PLSR+GRNN model yielded high R2 performance deviation (RPD), low root mean squared error (RMSE) with values 0.91, 3.15, 1.98, respectively, training test 0.81, 2.25, 2.46, validation test. Compared PLSR model, PLSR+BPNN higher were more stable. Moreover, overcame issue overfitting, which occurs a single predict ratio. Therefore, Fractional promising spectral preprocessing technique improve models. conclude forest effectively estimated using combined field spectroradiometer data