High-frequency leaf area index (LAI) dataset is essential for vegetation dynamic monitoring and crop yield estimation. However, due to the negative impacts of land surface heterogeneity, current hectometric-resolution LAI products cannot satisfy the uncertainty requirement of LAI dataset in practice. Here, we proposed a method named Utilization of Sub-Pixel Information (USPI) that leverages fine-scale remote sensing data to improve the accuracy of hectometric-resolution LAI retrieval. Specifically, based on machine learning (ML) models trained by representative samples, we retrieved the USPI LAI from MODIS reflectance by incorporating the sub-pixel information from Sentinel-2 LAI estimates. The USPI LAI was comprehensively evaluated using 30-m LAI reference maps in three aspects: the performance of different ML models, the comparison with MODIS LAI products, and the potential correction of USPI LAI for clumping effect. Results showed that Gaussian Process Regression (GPR) model outperformed other ML models for deriving LAI estimates. Furthermore, USPI LAI exhibited better performance than MODIS LAI product, with bias, root mean square error (RMSE), and R2of -0.308, 0.593, and 0.826, respectively, especially for pixels contaminated by atmospheric conditions. Nevertheless, the underestimation of USPI LAI should be noted because the effective LAI provided by Sentinel-2 was involved in the GPR training process. Thus, it is necessary to introduce the accurate clumping index dataset for further improvement of USPI LAI retrievals. Our study indicates that incorporating the sub-pixel information from decametric-resolution data can effectively reduce the uncertainty of hectometric-resolution LAI retrieval, which is promising for generating the high-accuracy LAI time series dataset.

Leaf area index (LAI) is a pivotal biophysical parameter for characterizing canopy structure and monitoring vegetation growth. Although the look-up table (LUT) method has been widely employed for LAI retrieval, the optimization of key retrieval processes remains to be explored. Here, we proposed a generic optimization strategy for LUT-based inversion based on Landsat -8 imagery and global ground LAI measurements. Specifically, based on the LUT generated by the PROSAIL model, LAI inversion was optimized by introducing several functions, including band selection, artificial noise addition, cost function (CF) substitution, and multiple solutions. Furthermore, the optimized LUT-based inversion method was compared to the Simplified Level 2 Product Prototype Processor (SL2P) method and the ground-measurement-derived (GMD) regression method to comprehensively evaluate its performance over various vegetation types. Results showed that the combination of Red, near-infrared (NIR), and shortwave infrared-1 (SWIR1) bands was well suited to capture LAI dynamics. In terms of accuracy and efficiency, the best performance was achieved by the optimal band combination and retrieval parameter settings (i.e., root-mean-square error (RMSE) as CF, noise level of 20%, and multiple solutions of 5%), with the RMSE and${R} ^{2}$of 0.817 and 0.740, respectively. In addition, the optimized LUT-based inversion was superior to SL2P method in accuracy and to GMD regression method in efficiency. Overall, the optimized LUT-based inversion strategy can be applied for estimating decametric-resolution LAI with high accuracy over different regions and observation dates at a global scale, exhibiting high adaptability and generalization capability, especially for crops, and requiring no ground LAI measurements.

High spatiotemporal resolution time series of leaf area index (LAI) are essential for monitoring crop dynamics and validating coarse-resolution LAI products. The optical satellite sensors at decametric-resolution have historically suffered from a long revisit cycle and cloud contamination issues that hampered the acquisition of frequent and high-quality observations. The 16-m/4-day resolution of the new generation Gaofen-1 (GF-1) and Gaofen-6 (GF-6) satellites provide an unprecedented opportunity to address these limitations. Here we developed an effective strategy to generate daily 16-m LAI maps combing GF-1/6 data and ground LAINet measurements. All high-quality GF-1/6 observations were utilized first to derive smoothed time series of vegetation indices (VIs). Second, a random forest regression (RF-r) model was trained to link the VIs with corresponding field LAI measurements. The trained RF-r was finally employed to generate the LAI maps. Results demonstrated the reliability of the reconstructed daily VIs (relative error2of 0.05, 0.59 and 0.75, respectively. The LAI time series well captured the spatiotemporal variation of crop growth. Furthermore, the continuous GF-1/6 LAI maps outperformed Sentinel-2 LAI estimates both in terms of temporal frequency and accuracy. Our study indicates the potential of GF-1/6 to generate continuous decametric-resolution LAI maps for fine-scale agricultural monitoring.

Leaf area index (LAI) is an important indicator for monitoring vegetation growth and estimating crop yields. The empirical-based model using vegetation indices (VIs) is an effective method for LAI estimation at the regional scale. However, due to the complexity of canopy radiation interaction processes, the leaf chlorophyll content (Cab) and saturation effects on canopy reflectance restrict the accuracy of VI-based LAI retrieval. To address these limitations, we propose a novel chlorophyll-insensitive vegetation index (CIVI) using red, red-edge and near-infrared bands to improve regional LAI mapping. The CIVI was developed based on the sensitivity analysis of red-edge band reflectance to LAI andCabusing the simulation dataset from the PROSAIL model. Then, the performance of CIVI was carefully evaluated from two aspects: the sensitivity of VI to LAI and other parameters, and the accuracy of LAI estimates using different VIs over homogeneous (cropland and grassland) and non-homogeneous (forest) biome canopies. The results suggested that CIVI can capture LAI variations well while remaining insensitive toCabvariations. Additionally, the sensitivity of CIVI to other vegetation biochemical and biophysical parameters did not increase significantly compared to that of other VIs. Furthermore, CIVI exhibited the best performance of LAI retrievals over both homogeneous (R2=0.938, RMSE=0.447 and rRMSE=21.3%) and non-homogenous (R2=0.635, RMSE=0.693 and rRMSE=14.0%) canopies among all selected VIs, especially for the high LAI. Our results indicated that the developed CIVI incorporating red-edge bands with a suitable formula can effectively reduce theCaband saturation effects, which is promising for improving VI-based LAI estimation.

Accurate estimation of photosynthetic phenology is of great importance for understanding carbon cycles. Most vegetation indices (VIs) calculated from remotely sensed reflectances represent the canopy structure and have high uncertainty in detecting the photosynthetic phenology. We compared the start/end of the photosynthetically active season (SOS/EOS) extracted from the normalized difference vegetation index (NDVI), the enhanced vegetation index (EVI), the near-infrared reflectance of vegetation (NIRv) and the product of NIRv and solar incident radiation (NIRvP) over northern deciduous broadleaf forests, and we used the metrics generated from solar-induced chlorophyll fluorescence (SIF), a proxy for photosynthesis, as reference. We found that the growing season extracted from the structural VIs was generally longer than the duration of photosynthetic activity retrieved from SIF: SOS derived from NDVI < NIRvP < EVI ≈ NIRv ≈ SIF and EOS from NDVI > NIRv ≈ EVI > NIRvP ≈ SIF. We investigated the mechanism underlying these phenological discrepancies using the paradigm of light-use efficiency. Our results show that the divergent performances of VIs were related to main factors limiting photosynthesis, which vary across different growth stages. The fraction of absorbed photosynthetically active radiation absorbed by chlorophyll (FAPARchl) that is well characterized by both EVI and NIRv, was the dominant factor of spring photosynthetic phenology, whilst NIRvP that is a proxy of the total amount of photosynthetically active radiation absorbed by chlorophyll (APARchl) was the dominant factor in autumn when radiation determines photosynthetic phenology. As such, we suggest that these factors be accounted for when selecting VIs for the extraction of photosynthetic phenology, i.e., EVI and NIRv are more suitable for accurate retrieval of SOS, and NIRvP is more suitable for accurate retrieval of EOS.

Vegetation index (VI) derived from remotely sensed images is a proxy of terrestrial vegetation information and widely used in land monitoring and global change studies. Recently, the prediction of vegetation properties has been an interest in related communities. With the accumulation of satellite records over the past few decades, the spatial–temporal prediction of VI becomes feasible. In this letter, we developed deep recurrent neural networks (RNNs) with long short-term memory (LSTM) and gated recurrent units (GRUs) to predict the short-term VI based on historical observations. The pixel-based fully connected networks GRU and LSTM (FCGRU and FCLSTM) and patch-based convolutional networks (ConvGRU and ConvLSTM) are established and compared with the traditional multilayer perceptron (MLP) model. Moderate Resolution Imaging Spectroradiometer (MODIS) and Sentinel-2 normalized difference VI (NDVI) data sets were used in the experiments. The prediction performance is evaluated globally in different regions, different vegetation types, and different growing seasons. Results demonstrate that the RNN models can predict VI with high accuracy (average root mean square error (RMSE) around 0.03), which is superior to the MLP model. In general, the pixel-based RNN models performed better than the patch-based models especially in regions with a larger proportion of outliers. And the prediction accuracy is stable over different vegetation types and growing seasons.

While EVI and LAI are the most widely used vegetation parameters which can be used for remote sensing phenology extraction, this paper aims at assessing the differences of phenology information extracted from EVI and LAI time series and exploring either EVI or LAI time series performs well for all vegetation types over a large scale. To achieve this, GLASS-LAI phenology product(GLP) was generated by the same algorithm with MODIS-EVI phenology product(MLCD) over China from 2001 to 2012. The two phenology products were compared in different climate regions and vegetation types over a large scale and evaluated by ground observations. Results show that the missing rate of GLP(11.90%) is less than that of MLCD(22.84%). The difference between GPL and MLCD varies in different climate regions and vegetation types. GLP performs better than MLCD in croplands and forests, while MLCD performs better than GLP in grasslands.

Surface upward longwave radiation (SULR) is a significant component of the surface radiation budget and is closely linked with evapotranspiration, soil moisture, and surface cooling on clear nights. Therefore, accurately estimating SULR is essential to better understand its spatiotemporal dynamics or to characterize the thermal environment of a given land surface. Currently, most methods for estimating SULR (including the physical and hybrid methods) fail to account for the thermal anisotropy, which can introduce significant errors into the calculation. We previously proposed the combined algorithm that considers the thermal anisotropy to more accurately estimate the SULR. However, this proposed method has several shortcomings. For example, it considers the directionality of the emissivity and the effective temperature separately under the support of a parametric directional emissivity model. However, the directional emissivity model is not maturely developed for different land surface types, especially on non-vegetated surfaces. And the separation of land surface temperature and emissivity may undermine the estimation accuracy. Furthermore, this proposed method requires a series of input parameters that is not always available, limiting its applicability. In this paper, we present a refined algorithm that uses a kernel-driven model and the technique of band conversion to calculate the SULR directly based on surface-leaving radiances. This direct physical algorithm is then applied to the Wide-angle infrared Dual-mode line/area Array Scanner data set and validated using longwave radiation data collected by automatic meteorological stations from the Heihe Watershed Allied Telemetry Experimental Research experiment. The results of these tests suggest that the direct algorithm works effectively. The root-mean-square error (RMSE) and mean bias error (MBE) of the direct algorithm on maize surfaces are 4.417 and 0.474 W · m-2, respectively. When the thermal anisotropy is incorporated, the RMSE and absolute MBE decrease by a maximum of 4.734 and 7.414 W·m-2, respectively. Different land types yield different results: for vegetable surfaces, the estimation biases of the direct model are approximately -2 W · m-2, whereas orchard surfaces yield biases are between -2 and -3.5 W · m-2, and village surfaces yield biases exceeding -10 W · m-2. These differences can be attributed to the varying effects of the kernel-driven model across different types of land surfaces. The RMSE and absolute MBE obtained using the direct algorithm are slightly smaller (0.587 and 1.685 W·m-2, respectively) than those obtained using the combined algorithm; they are also smaller than the results of the traditional temperature-emissivity algorithm (by 8.7 and 11.7 W · m-2, respectively).