Weiliang Fan

Estimation of aboveground carbon stock of Moso bamboo (Phyllostachys heterocycla var. pubescens) forest with a Landsat Thematic Mapper image

The extensive distribution of bamboo forests in South and Southeast Asia plays an important role in the global carbon budget. It is an urgent task to accurately and in good time estimate carbon stock within these areas. In this study, linear regression, partial least-squares (PLS) regression and backpropagation artificial neural network (BP-ANN) with a Gaussian error function as the activation function of the hidden layers (Erf-BP) were used to estimate aboveground carbon (AGC) stock of Moso bamboo in Anji, Zhejiang Province, China. Based on the combined use of Landsat Thematic Mapper (TM) and field measurements, the results indicate that the Erf-BP model provided the best estimation performance, and the linear regression model performed the poorest. This study indicates that remote sensing is an effective way of estimating AGC of Moso bamboo in a large area.

GOST: A Geometric-Optical Model for Sloping Terrains

GOST is a geometric-optical (GO) model for sloping terrains developed in this study based on the four-scale GO model, which simulates the bidirectional reflectance distribution function (BRDF) of forest canopies on flat surfaces. The four-scale GO model considers four scales of canopy architecture: tree groups, tree crowns, branches, and shoots. In order to make this model suitable for sloping terrains, the mathematical description for the projection of tree crowns on the ground has been modified to consider the fact that trees grow vertically rather than perpendicularly to sloping grounds. The simulated canopy gap fraction and the area ratios of the four scene components (sunlit foliage, sunlit background, shaded foliage, and shaded background) by GOST compare well with those simulated by 3-D virtual canopy computer modeling techniques for a hypothetical forest. GOST simulations show that the differences in area ratios of the four scene components between flat and sloping terrains can reach up to 50%-60% in the principal plane and about 30% in the perpendicular plane. Two case studies are conducted to compare modeled canopy reflectance with observations. One comparison is made against Landsat-5 Thematic Mapper (TM) reflectance, demonstrating the ability of GOST to model canopy reflectance variations with slope and aspect of the terrain. Another comparison is made against MODIS surface reflectance, showing that GOST with topographic consideration outperforms that without topographic consideration. These comparisons confirm the ability of GOST to model canopy reflectance on sloping terrains over a large range of view angles.

Regional Leaf Area Index Retrieval Based on Remote Sensing: The Role of Radiative Transfer Model Selection

Physically-based approaches for estimating Leaf Area Index (LAI) using remote sensing data rely on radiative transfer (RT) models. Currently, many RT models are freely available, but determining the appropriate RT model for LAI retrieval is still problematic. This study aims to evaluate the necessity of RT model selection for LAI retrieval and to propose a retrieval methodology using different RT models for different vegetation types. Both actual experimental observations and RT model simulations were used to conduct the evaluation. Each of them includes needleleaf forests and croplands, which have contrasting structural attributes. The scattering from arbitrarily inclined leaves (SAIL) model and the four-scale model, which are 1D and 3D RT models, respectively, were used to simulate the synthetic test datasets. The experimental test dataset was established through two field campaigns conducted in the Heihe River Basin. The results show that the realistic representation of canopy structure in RT models is very important for LAI retrieval. If an unsuitable RT model is used, then the root mean squared error (RMSE) will increase from 0.43 to 0.60 in croplands and from 0.52 to 0.63 in forests. In addition, an RT model’s potential to retrieve LAI is limited by the availability of a priori information on RT model parameters. 3D RT models require more a priori information, which makes them have poorer generalization capability than 1D models. Therefore, physically-based retrieval algorithms should embed more than one RT model to account for the availability of a priori information and variations in structural attributes among different vegetation types.

Satellite-based carbon stock estimation for bamboo forest with a non-linear partial least square regression technique

This article explores a non-linear partial least square (NLPLS) regression method for bamboo forest carbon stock estimation based on Landsat Thematic Mapper (TM) data. Two schemes, leave-one-out (LOO) cross validation (scheme 1) and split sample validation (scheme 2), are used to build models. For each scheme, the NLPLS model is compared to a linear partial least square (LPLS) regression model and multivariant linear model based on ordinary least square (LOLS). This research indicates that an optimized NLPLS regression mode can substantially improve the estimation accuracy of Moso bamboo (Phyllostachys heterocycla var. pubescens) carbon stock, and it provides a new method for estimating biophysical variables by using remotely sensed data.

Hybrid Geometric Optical–Radiative Transfer Model Suitable for Forests on Slopes

A new geometric optical (GO)-radiative transfer (RT) model with a multiple scattering scheme suitable for sloping forest canopies is developed in this study. It is based on a Geometrical-Optical model for Sloping Terrains and an RT method. This new model overcomes the difficulty to prescribe bidirectional reflectance factors (BRFs) of shaded components (shaded foliage and background) in GO modeling through simulating radiation multiple scattering within a sloping forest. A case study shows that multiply scattered radiation depends on topographic factors and leaf area index. The contributions of the shaded components to stand-level BRF are less than 3% in the red band and can reach up to 40% in the near-infrared (NIR) band. The “multiangle” Moderate Resolution Imaging Spectroradiometer (MODIS) data over sloping pixels are selected to validate the modeled forest BRF. Considering the multiple scattering schemes and topographic factors, the modeled BRF is closer to the MODIS surface reflectance (BRF product) (red band: R2 = 0.8614, rRMSE = 0.1339; NIR band: R2 = 0.7573, rRMSE = 0.0850) than the modeled BRF (red band: R2 = 0.7771, rRMSE=0.1839; NIR band: R2 =0.5176, rRMSE = 0.1155) without topographic consideration. It is also shown that the MODIS surface reflectance of sloping forests at multiple angles can be simulated well using the newly developed model.

Applicability of spectral and spatial information from IKONOS-2 imagery in retrieving leaf area index of forests in the urban area of Nanjing, China

Abstract. Remote sensing is currently an indispensable tool for retrieving the leaf area index (LAI) of forests. However, the applicability of remote sensing in retrieving LAI of forests in urban areas has not been thoroughly investigated. The ability of spectral and spatial information from IKONOS-2 imagery to retrieve LAI of forests was studied through analyzing the correlations of four commonly used vegetation indices (VIs) and four texture measures (TEXs) with LAI measured at different types of plots in the urban area of Nanjing, China and comparing the ability of models based on these parameters to estimate LAI of forests. The results show that VIs and TEXs calculated from the high-resolution remote sensing data are both applicable in retrieving LAI of forests in urban areas. The relative advantages of VIs and TEXs are related to the density and spatial regularity of forests. TEX exceeds VI for regularly planted low broad-leaf forests with low density owing to the deterioration of the linkage of VIs with canopy LAI caused by strong soil noise. For forests with moderate and high density, VI exceeds TEX in the retrieval of LAI. As to natural broad-leaf forests with high density and spatial complexity, combining VI and TEX can improve the accuracy of the retrieved LAI by 8.9% to 27.0%. VIs and TEXs are exclusive in retrieving LAI due to the intrinsic linkages of these parameters. The atmospherically resistant vegetation index over-perform other VIs in retrieving LAI of forests owing to its ability to constrain atmospheric disturbance on remote sensing data, which is serious and exhibits great spatial variability in the study area.

A Radiative Transfer Model for Heterogeneous Agro-Forestry Scenarios

Landscape heterogeneity is a common natural phenomenon but is seldom considered in current radiative transfer (RT) models for predicting the surface reflectance. This paper developed an analytical RT model for heterogeneous Agro-Forestry scenarios (RTAF) by dividing the scenario into nonboundary regions (NRs) and boundary regions (BRs). The scattering contribution of the NRs can be estimated from the scattering-by-arbitrarily-inclined-leaves-with-the-hot-spot-effect model as homogeneous canopies, whereas that of the BRs is calculated based on the bidirectional gap probability by considering the interactions and mutual shadowing effects among different patches. The multiangular airborne observations and discrete-anisotropic-RT model simulations were used to validate and evaluate the RTAF model over an agro-forestry scenario in the Heihe River Basin, China. The results suggest that the RTAF model can accurately simulate the hemispherical-directional reflectance factors (HDRFs) of the heterogeneous scenarios in the red and near-infrared (NIR) bands. The boundary effect can significantly influence the angular distribution of the HDRFs and consequently enlarge the HDRF variations between the backward and forward directions. Compared with the widely used dominant cover type (DCT) and spectral linear mixture (SLM) models, the RTAF model reduced the maximum relative error from 25.7% (SLM) and 23.0% (DCT) to 9.8% in the red band and from 19.6% (DCT) and 13.7% (SLM) to 8.7% in the NIR band. The RTAF model provides a promising way to improve the retrieval of biophysical parameters (e.g., leaf area index) from remote sensing data over heterogeneous agro-forestry scenarios.

Ability of the Photochemical Reflectance Index to Track Light Use Efficiency for a Sub-Tropical Planted Coniferous Forest

Light use efficiency (LUE) models are widely used to estimate gross primary productivity (GPP), a dominant component of the terrestrial carbon cycle. Their outputs are very sensitive to LUE. Proper determination of this parameter is a prerequisite for LUE models to simulate GPP at regional and global scales. This study was devoted to investigating the ability of the photochemical reflectance index (PRI) to track LUE variations for a sub-tropical planted coniferous forest in southern China using tower-based PRI and GPP measurements over the period from day 101 to 275 in 2013. Both half-hourly PRI and LUE exhibited detectable diurnal and seasonal variations, and decreased with increases of vapor pressure deficit (VPD), air temperature (Ta), and photosynthetically active radiation (PAR). Generally, PRI is able to capture diurnal and seasonal changes in LUE. However, correlations of PRI with LUE varied dramatically throughout the growing season. The correlation was the strongest (R2 = 0.6427, p 0.3) with moderate to high VPD (>20 hPa) and high temperatures (>31 C). Overall, we found that PRI is most sensitive to variations in LUE under stressed conditions, and the sensitivity decreases as the growing conditions become favorable when atmosphere water vapor, temperature and soil moisture are near the optimum conditions.

Retrieval of Canopy Closure and LAI of Moso Bamboo Forest Using Spectral Mixture Analysis Based on Real Scenario Simulation

This paper investigates the retrievals of the canopy closure and leaf area index (LAI) of the Moso bamboo forest from the Landsat Thematic Mapper data using a constrained linear spectral unmixing method. A new approach for endmember collection based on the real scenario simulation of the Moso bamboo forest is developed. Four fraction images (i.e., sunlit canopy, shaded canopy, sunlit background, and shaded background) are calculated and used to develop the canopy closure and LAI. The results show that the predicted crown closure, which was inverted from the sunlit and shaded canopies, has a good agreement with the observed crown closure (R2 = 0.725). The accuracy assessment indicates that the root mean square error (rmse) and the relative root mean square error (rmse_r) are 10% and 13.37% for the predicted crown closure, respectively. The LAI has the highest correlation coefficient with the shaded background, and it can be fitted by an exponential model (R2 = 0.497). The linear relationship between the predicted and observed LAI values is significant at a level of 99% (P <; 0.01 and R2 = 0.459), and the LAI can be predicted by the exponential model.

Topographic Correction of Forest Image Data Based on the Canopy Reflectance Model for Sloping Terrains in Multiple Forward Mode

Topographic correction methods rarely consider the canopy parameter effects directly and explicitly for sloping canopies. In order to address this problem, the topographic correction method MFM-GOST2 was developed by implementing the second version of the Geometric-Optical model for Sloping Terrains (the GOST2 model) in the multiple forward mode (MFM) inversion framework. First, a look up table (LUT) was constructed by multiple forward modeling of the GOST2 model; second, the radiance of a remotely sensed image and its corresponding topographic data were used for searching potential canopy parameter combinations from the LUT; and third, the corrected radiance was determined by averaging potential radiances of horizontal canopies from the LUT according to the canopy parameter combinations. The MFM-GOST2 and twelve generally used topographic correction methods were evaluated via a case study by visual analysis, linear relationship analysis, and the rose diagram analysis. The result showed that the MFM-GOST2 method successfully removed most of the topographic effects of a subset image of the Landsat-8 image in a case study. The case study also illustrates that the rose diagram analysis is a good way to evaluate topographic corrections, but the linear relationship analysis cannot be used independently for the evaluations because the decorrelation is not a sufficient condition to determine a successful topographic correction.