Search

Most of the current optical-based forest AGB estimation methods only use spatial information of reflectance, making estimates susceptible to saturation and noise. We introduced dynamic time-series information on forest reflectance into AGB estimation and explored its ability to improve estimation accuracy. The experiments were conducted based on 10304 in situ AGB plots and corresponding Landsat 5/7/8 data from 1984 to 2019 in Australia, using a random forest algorithm for AGB estimation. Results show that time-series modeling of reflectance could remove most of the noise in observed reflectance, and improved the correlation between reflectance and AGB. After introducing the time-series parameters into AGB estimation, estimation accuracy was significantly improved, with R2 increased by 0.08 and RMSE reduced by 7.5 tons/ha.

PolInSAR technique has been proved effective in mapping forest height. However, for P band repeat-pass PolInSAR data, the influence of residual ground scattering contribution and temporal decorrelation are non-negligible for forest height inversion. To resolve these two problems, we developed a dual-baseline inversion method to reduce the influence of ground scattering contribution, and then cooperated it with Random-Motion-over-Ground (RMoG) model to compensate for the influence of temporal decorrelation. The Three-stage inversion showed a severe overestimation for the whole image, producing R2 of ca. 0.4 and RMSE of ca. 7m. For near range (with $k_{z}$ larger than 0.1rad/m), the overestimation was mainly caused by ground scattering contribution, which was effectively solved by using fixed extinction method. For far range (with $k_{z}$ smaller than 0.06rad/m), the influence of ground scattering contribution seemed decreased because the fixed extinction method did not reduce the overestimation, and overestimation was mainly caused by volume temporal decorrelation. Besides, the influence of ground scattering contribution on forest height inversion seemed greater than that of temporal decorrelation. The proposed method successfully solved both of the two influences, and reduced the overestimation for both near and far range, producing the best accuracy with R2 of 0.62 and RMSE of 3.73m.

Fuel load is the key factor driving fire ignition, spread and intensity. The current literature reports the light detection and ranging (LiDAR), optical and airborne synthetic aperture radar (SAR) data for fuel load estimation, but the optical and SAR data are generally individually explored. Optical and SAR data are expected to be sensitive to different types of fuel loads because of their different imaging mechanisms. Optical data mainly captures the characteristics of leaf and forest canopy, while the latter is more sensitive to forest vertical structures due to its strong penetrability. This study aims to explore the performance of Landsat Enhanced Thematic Mapper Plus (ETM+) and Advanced Land Observing Satellite (ALOS) Phased Arrayed L-band Synthetic Aperture Radar (PALSAR) data as well as their combination on estimating three different types of fuel load—stem fuel load (SFL), branch fuel load (BFL) and foliage fuel load (FFL). We first analyzed the correlation between the three types of fuel load and optical and SAR data. Then, the partial least squares regression (PLSR) was used to build the fuel load estimation models based on the fuel load measurements from Vindeln, Sweden, and variables derived from optical and SAR data. Based on the leave-one-out cross-validation (LOOCV) method, results show that L-band SAR data performed well on all three types of fuel load (R2 = 0.72, 0.70, 0.72). The optical data performed best for FFL estimation (R2 = 0.66), followed by BFL (R2 = 0.56) and SFL (R2 = 0.37). Further improvements were found for the SFL, BFL and FFL estimation when integrating optical and SAR data (R2 = 0.76, 0.81, 0.82), highlighting the importance of data selection and combination for fuel load estimation.

We developed a new algorithm called MFmask (Mountainous Fmask) for automated cloud and cloud shadow detection for Landsats 4–8 images acquired in mountainous areas. The MFmask algorithm, built upon the success of the Fmask algorithm (Zhu and Woodcock, 2012; Zhu et al., 2015), is designed for cloud and cloud shadow detection in mountainous areas, where the Fmask algorithm is not performing well. The inputs of the MFmask algorithm include Landsat Top of Atmosphere (TOA) reflectance, Brightness Temperature (BT), and Digital Elevation Models (DEMs). Compared to Fmask, MFmask can separate water and land pixels better in mountainous areas with the aid of DEMs. Moreover, MFmask produces better cloud detection results than Fmask in mountainous areas after BT is linearly normalized by DEMs. To provide more accurate cloud shadow detection in mountainous areas, MFmask uses a double-projection approach to better predict cloud shadow shape on slope side. Additionally, MFmask applies a topographic correction to remove terrain shadows and estimates cloud base height with neighboring clouds. Both will reduce the possibility of cloud and cloud shadow mismatch and increase cloud shadow detection accuracy for places with large topographic gradient. To test the performance of the proposed MFmask algorithm, a total of 67 Landsat images acquired in mountainous areas from different parts of the world were selected for assessing the accuracy of cloud detection, in which 15 of them were used for assessing the accuracy of cloud shadow detection. Compared with Fmask, MFmask can provide substantial improvements in cloud and cloud shadow detection accuracies for places with large topographic gradient and also work well for relatively flat terrain.

Forest fuel moisture content (FMC) dynamics are paramount to assessing the forest wildfire risk and its behavior. This variable can be retrieved from remotely sensed data using a radiative transfer model (RTM). However, previous studies generally treated the background of forest canopy as soil surface while ignored the fact that the soil may be covered by grass canopy. In this study, we focused on retrieving FMC of such forestry structure by coupling two RTMs: PROSAIL and PRO-GeoSail. The spectra of lower grass canopy were firstly simulated by the PROSAIL model, which was then coupled into the PRO-GeoSail model. The results showed that the accuracy level of retrieved FMC using this coupled model was better than that when the PRO-GeoSail model used alone. Further analysis revealed that low FMC condition fostered by fire weather condition had an important influence on the breakout of a fire during the study period.

This paper presents a novel method to derive grassland aboveground biomass (AGB) based on the PROSAILH (PROSPECT + SAILH) radiative transfer model (RTM). Two variables, leaf area index (LAI, m2m−2, defined as a one-side leaf area per unit of horizontal ground area) and dry matter content (DMC, gcm−2, defined as the dry matter per leaf area), were retrieved using PROSAILH and reflectance data from Landsat 8 OLI product. The result of LAI × DMC was regarded as the estimated grassland AGB according to their definitions. The well-known ill-posed inversion problem when inverting PROSAILH was alleviated using ecological criteria to constrain the simulation scenario and therefore the number of simulated spectra. A case study of the presented method was applied to a plateau grassland in China to estimate its AGB. The results were compared to those obtained using an exponential regression, a partial least squares regression (PLSR) and an artificial neural networks (ANN). The RTM-based method offered higher accuracy (R2 = 0.64 and RMSE = 42.67 gm−2) than the exponential regression (R2 = 0.48 and RMSE = 41.65 gm−2) and the ANN (R2 = 0.43 and RMSE = 46.26 gm−2). However, the proposed method offered similar performance than PLSR as presented better determination coefficient than PLSR (R2 = 0.55) but higher RMSE (RMSE = 37.79 gm−2). Although it is still necessary to test these methodologies in other areas, the RTM-based method offers greater robustness and reproducibility to estimate grassland AGB at large scale without the need to collect field measurements and therefore is considered the most promising methodology.

In this study, the burn severity of several wildfires ignited at northern Australian tropical savannas area were estimated using the Forest Reflectance and Transmittance (FRT) radiative transfer model (RTM) and Sentinel-2A Multi-Spectral Instrument (MSI) satellite data. To alleviate the spectral confusion between severe (SV) and not-severe (NSV) burnt levels caused by sparse tree distribution, the MODIS Vegetation Continuous Fields (VCF) tree cover percentage data was used to constrain the inversion. The results showed that the accuracy of burn severity estimation significantly improves when considering the tree coverage, with overall accuracy for two study sites increasing from 65% to 81% and kappa coefficient from 0.35 to 0.55. Future work will focus on extending the methodology to other ecosystems.

With the upcoming BIOMASS mission, P-band PolInSAR is expected to provide new perspectives on global forest aboveground biomass (AGB). However, its performance has not yet been fully evaluated for dense tropical forests with complex structure and very high biomass. Based on the TropiSAR campaign in French Guiana, we explored the challenges of the three most commonly used PolInSAR measures to capture AGB in tropical forests; coherence magnitude, interferometric phase, and backscatter. An improved AGB estimation approach was developed by integrating multiple information derived from single-baseline PolInSAR data. The approach involves ground-volume backscatter decomposition and combines volume backscatter with the retrieved forest height. Volume backscatter from the forest canopy was the best predictor of AGB for tropical forests, whereas the ground backscatter contribution was affected by the complex underlying surface and terrain slope. Both LiDAR- and PolInSAR-derived forest heights showed limited correlation with high AGB due to the varying forest basal area. The linear combination of PolInSAR-derived forest height and volume backscatter complemented each other and produced improved AGB estimates. Comparing three different PolInSAR data pairs, the proposed method produced an AGB map with an average R2 of 0.7 and RMSE of 34 tons/ha (relative RMSE of 9.4%) at a spatial resolution of 125 × 125 m2 for biomass between 250–500 tons/ha.

Live fuel moisture content (LFMC) is an important factor in wildfire risk assessment, but the estimation of LFMC is affected by the leaf area index (LAI) if it is retrieved using a radiative transfer model. To reduce the influence of LAI in the LFMC retrieval process, we used the interval estimated LAI to parameterize the PROSAIL ( ${\mathbf{PROSPECT}} + {\mathbf {SAIL}}$ ) radiative transfer model. The LAI values were first estimated based on the MODIS LAI product (MOD15A2) using a simple downscaling technique, and then were constrained using an interval estimation method at the confidence levels of 80%, 90%, and 95%. To evaluate the reliability of this approach, another two LAI constraint strategies were applied, weakly constrained LAI over a wide range and strongly constrained LAI with fixed values. A look up table (LUT) was set up for the parameterized PROSAIL model. Furthermore, the LFMC can be retrieved based on the Landsat TM and 8 products. The results showed that the retrieved LFMC using the interval estimated LAI at the 95% confidence level resulted in the best accuracy level ( ${\mathbf{R}}^{\mathbf{2}} = {\mathbf{0.78}}$ , ${\mathbf{RMSE}} = {\mathbf{35.80}}\% $ ), which was slightly better than the LFMC obtained when using the interval estimated LAI at confidence levels of 90% ( ${\mathbf{R}}^{\mathbf{2}} = {\mathbf{0.75}},\;{\mathbf{RMSE}} = {\mathbf{39.38}}\% $ ) and 80% ( ${\mathbf{R}}^{\mathbf{2}} = {\mathbf{0.70}},\;{\mathbf{RMSE}} = {\mathbf{41.21}}\% $ ). The accuracy levels of the retrieved LFMC using weakly constrained LAI ( ${\mathbf{R}}^{\mathbf{2}} = {\mathbf{0.64}},\;{\mathbf{RMSE}} = {\mathbf{41.68}}\% $ ) and strongly constrained LAI ( ${\mathbf{R}}^{\mathbf{2}} = {\mathbf{0.68}},\;{\mathbf{RMSE}} = {\mathbf{45.53}}\% $ ) were poor. Thus, this study demonstrated that LFMC retrieval could be improved by using the interval estimated LAI.

In this study, an arid grassland was selected, and the chlorophyll content of the leaf and canopy level was estimated based on Landsat-8 Operational Land Imager OLI data using the PROSAIL radiative transfer RT model. Two vegetation indices green chlorophyll index, CIgreen, and greenness index, G were selected to estimate the leaf and canopy chlorophyll content LCC and CCC. By analysing the effect of soil background on the two indices, the LCC was divided into low and moderate-to-high levels. A different combination of the two indices was adopted at each level to improve the chlorophyll content estimation accuracy. The results suggested that the chlorophyll content estimated using the proposed method yielded a higher accuracy with coefficient of determination, R2 = 0.84, root-mean-square error, RMSE = 9.67 μg cm−2 for LCC and R2 = 0.85, RMSE = 0.43 g m−2 for CCC than that using CIgreen alone with R2 = 0.62, RMSE = 20.04 μg cm−2 for LCC and R2 = 0.85, RMSE = 0.71 g m−2 for CCC. The results also confirmed the validity of this approach to estimate the chlorophyll content in arid areas.

Synthetic Aperture Radar (SAR) texture has been demonstrated to have the potential to improve forest biomass estimation using backscatter. However, forests are 3D objects with a vertical structure. The strong penetration of SAR signals means that each pixel contains the contributions of all the scatterers inside the forest canopy, especially for the P-band. Consequently, the traditional texture derived from SAR images is affected by forest vertical heterogeneity, although the influence on texture-based biomass estimation has not yet been explicitly explored. To separate and explore the influence of forest vertical heterogeneity, we introduced the SAR tomography technique into the traditional texture analysis, aiming to explore whether TomoSAR could improve the performance of texture-based aboveground biomass (AGB) estimation and whether texture plus tomographic backscatter could further improve the TomoSAR-based AGB estimation. Based on the P-band TomoSAR dataset from TropiSAR 2009 at two different sites, the results show that ground backscatter variance dominated the texture features of the original SAR image and reduced the biomass estimation accuracy. The texture from upper vegetation layers presented a stronger correlation with forest biomass. Texture successfully improved tomographic backscatter-based biomass estimation, and the texture from upper vegetation layers made AGB models much more transferable between different sites. In addition, the correlation between texture indices varied greatly among different tomographic heights. The texture from the 10 to 30 m layers was able to provide more independent information than the other layers and the original images, which helped to improve the backscatter-based AGB estimation.

Burn severity mapping greatly informs fire management and can be used to predict post-fire vegetation recovery. Satellite remote sensing is a cost-effective method for estimating burn severity, providing a comprehensive spatially explicit view of whole landscapes. However, the proportion of tree canopy cover (TCC) affects the reflectance signal, obscuring background char and ash. Consequently, traditional optical satellite remote sensing methods that do not account for variation in TCC misclassify burn severity, especially in areas with extremely low or high TCC. In this study, TCC data served to parameterize and constrain the inversion of the Forest Reflectance and Transmittance (FRT) radiative transfer model (RTM) to alleviate spectral confusion when retrieving burn severity. The methodology was evaluated using field measurements of burn severity for a series of wildfires in the fire-prone tropical savannas of northern Australia and the western United States. Burn severity classes were used for Australia while the Composite Burn Index (CBI) for US. Reflectance data from Sentinel-2A Multi-Spectral Instrument (MSI) and Landsat-5 Thematic Mapper (TM) corresponding to post-fire field survey dates were used to retrieve burn severity using FRT RTM (with and without using TCC information in its parameterization and inversion) and two standard empirical burn indices, dNBR and RdNBR, for comparison. Using FRT RTM without TCC constraint produced an overestimation for low burn severity in regions with low TCC and an underestimation for moderate and high burn severity in regions with high TCC. Burn severity estimation accuracy significantly improved by integrating TCC in the parameterization and inversion of FRT RTM. The overall accuracy in northern Australia increased from 65% to 81%, and the kappa coefficient increased from 0.35 to 0.55. In the western United States, R2 between estimated and observed CBI, increased from 0.33 to 0.54, root mean square error (RMSE) reduced from 0.53 to 0.43, and in all instances, the method performed better than dNBR and RdNBR. The method used in this study achieved more accurate burn severity mapping, thus assisting land managers to better understand post-fire vegetation resilience and forest management.

The Multi Spectral Instrument (MSI) onboard Sentinel-2 can record the information in Vegetation Red-Edge (VRE) spectral domains. In this study, the performance of the VRE bands on improving land cover classification was evaluated based on a Sentinel-2A MSI image in East Texas, USA. Two classification scenarios were designed by excluding and including the VRE bands. A Random Forest (RF) classifier was used to generate land cover maps and evaluate the contributions of different spectral bands. The combination of VRE bands increased the overall classification accuracy by 1.40 %, which was statistically significant. Both confusion matrices and land cover maps indicated that the most beneficial increase was from vegetation-related land cover types, especially agriculture. Comparison of the relative importance of each band showed that the most beneficial VRE bands were Band 5 and Band 6. These results demonstrated the value of VRE bands for land cover classification.

Polarimetric Synthetic Aperture Radar Interferometry (PolInSAR) has shown potential for the retrieval of a forest canopy height model (CHM) and the underlying solid earth digital terrain model (DTM). However, because of non-volume decorrelation and other unavoidable errors, the robustness of retrieval heights is sensitive to the spatial baseline of the selected InSAR pairs, which relates forest parameters to measured coherence. Within the context of the random volume over ground (RVoG) model and the three-stage inversion method, we aimed to quantify the influence of spatial baseline on the inversions at P-band, which are distinct from the inversions at higher frequency due to the non-negligible ground contributions. This information assists in optimal baseline selection and the development of robust inversion schemes. Assumptions about the extinction coefficient and additional DTM or DEM were used to reduce the influence of ground contribution on CHM and DTM inversion, respectively. Inversions from published airborne repeat-pass P-band PolInSAR data with four different spatial baselines were validated against LiDAR-derived DTM and CHM data. The results show that a longer spatial baseline performed better in DTM inversion. The longest baseline produced the best R2 of 0.995 and RMSE of 0.555 m, much better than the smallest baseline with an R2 of 0.794 and RMSE of 3.74 m. A threshold height could be identified that determines the overestimation and underestimation of CHM inversion due to the non-volume decorrelation. Different baselines produced different threshold heights, making CHM inversion only accurate for a limited range of forest height around the threshold. The optimal baseline produced a CHM with R2 of 0.605 and RMSE of 2.67 m. Additionally, we found that using multiple baselines has the potential to improve CHM inversion, improving the R2 to 0.827 and RMSE to 0.876 m in our study.

Existing drought indices have been widely used to monitor meteorological drought and agricultural drought, however, few of them are focus on drought monitoring for grassland regions. This study presented a new drought index, the Grassland Drought Index (GDI), for monitoring drought conditions in global grassland regions. These regions are vital for the environment and human society but susceptible to drought. The GDI was constructed based on three measures of water content: precipitation, soil moisture (SM), and canopy water content (CWC). The precipitation information was extracted from the available precipitation datasets, and SM was estimated by downscaling exiting soil moisture data to a 1 km resolution, and CWC was retrieved based on the PROSAIL (PROSPECT + SAIL) model. Each variable was scaled from 0 to 1 for each pixel based on absolute minimum and maximum values over time, and these scaled variables were combined with the selected weights to construct the GDI. According to validation at the regional scale, the GDI was correlated with the Standardized Precipitation Index (SPI) to some extent, and captured most of the drought area identified by the United States Drought Monitor (USDM) maps. In addition, the global GDI product at a 1 km spatial resolution substantially agreed with the global Standardized Precipitation Evapotranspiration Index (SPEI) product throughout the period 2005–2010, and it provided detailed and accurate information about the location and the duration of drought based on the evaluation using the known drought events.

In this study, the Standardized Precipitation Evaporation Index (SPEI) was applied to characterize the drought conditions in Southwest China from 1982–2012. The SPEI was calculated by precipitation and temperature data for various accumulation periods. Based on the SPEI, the multi-scale patterns, the trend, and the spatio-temporal extent of drought were evaluated, respectively. The results explicitly showed a drying trend of Southwest China. The mean SPEI values at five time scales all decreased significantly. Some moderate and severe droughts were captured after 2005 and the droughts were even getting aggravated. By examining the spatio-temporal extent, the aggravating condition of drought was further revealed. To investigate the performance of SPEI, correlation analysis was conducted between SPEI and two remotely sensed drought indices: Soil Moisture Condition Index (SMCI) and Vegetation Condition Index (VCI). The comparison was also conducted with the Standardized Precipitation Index (SPI). The results showed that for both SMCI and VCI, the SPI and SPEI had approximate correlations with them. The SPEI could better monitor the soil moisture than the SPI in months with significant increase of temperature. The correlations between the VCI and SPI/SPEI were lower; nevertheless, the SPEI was slightly superior to the SPI.

Two-dimensional GIS are extensively applied to mineral potential mapping on a regional scale. However, these systems are unable to represent underground geological information in three spatial dimensions. The objective of this article is to overcome this defect and to develop a prototype system to qualitatively and quantitatively analyze underground mineral resources on the local scale, based on spatial data mining methods and three-dimensional GIS. The presented approach is based on 3D geological model established by the three-dimensional geological modeling software, such as Micromine, and is characterized by three-dimensional visualization, data management, and functionality for mineral resources prospectivity and quantitative assessment integrating weights-of-evidence model and case-based reasoning. The resulting system enables geologists to more accurately analyze the prospectivity and reserves of unknown mineral resource on the local scale.

Under the indoor building environment, when the fires and other accidents occur, how to effectively organize the masses evacuation and fire rescue, is closely related to the safety of people's lives and property and has become a critical problem of public concern. This paper presents an improved ant colony algorithm (ACO) to solve the problem of how to optimize the evacuation route and rescue route when an accident occurs. According to the key factors affecting people emergency evacuation, such as indoor building environment, fire and its combustion products, problem of path's optimal selection, etc., we propose an emergency evacuation model, based on the model it can give an optimal evacuation route for the mass and an optimal rescue route for the firefighters. We also analyzes the search results, it shows that the search results is robust and reasonable.

Grassland, which has a great significance to human society, is susceptible to the drought. An efficient way to monitor the drought scale of grassland is imperative. In this paper, we proposed a new drought index named Grassland Drought Index (GDI) to monitor the drought conditions of the grassland ecosystem. The GDI was constructed by comprehensively integrating the soil moisture content, canopy water content (CWC) and precipitation together with different weights. The soil moisture with 1km spatial resolution was estimated by downscaling the AMSR-E soil moisture from 25km to 1km. The CWC was retrieved by the 1km MODIS products based on the PROSAIL model. And the precipitation was acquired by CRU data. Besides, the global distribution product of GDI in July 2010 was produced. By validation, the correlation coefficient R between the scaled GDI and the Standard Precipitation Index (SPI) was 0.6109, and the root-mean-square error (RMSE) was 0.7774.

Understanding the dynamics of vegetation activity and its interaction with climate factors becomes an important issue in global change research. In this study, using satellite-derived normalized difference vegetation index (NDVI) data, we analyzed changes in vegetation growth at the region scale from 2000 to 2013 in Sichuan Province, China. And then explored the relationship between NDVI trends and the temperature and precipitation. The results showed: (1) The NDVI of the vegetated area in Sichuan was decreasing with a trend about −2.3 ×10−3 NDVI.yr−1. About 2/3 of the vegetated area showed a negative trend of annual mean NDVI. (2) In larger areas vegetation growth was controlled by temperature. (3) Monthly NDVI was significantly correlated with the preceding month's temperature and precipitation. (4) The spatial pattern of relationship between NDVI and climatic factors was likely correlated with terrain and vegetation type.

The retrieval of parameters through a physical mechanism model is promising for its generality but is challenged by the ill-posed inversion problem. This study focused on the use of multiple priori information to alleviate the ill-posed inversion problem. The priori information included the products of satellite images, the correlations among model free parameters, field survey, and the achievements of previous studies. However, the priori information was of uncertainty, which was described by multi-variables probability distribution in this study. A Bayesian network algorithm was used to retrieve the canopy water content (CWC) by calculating the posterior probability distribution of CWC based on the priori information, the HJ-1B product, and the PROSAIL model. The retrieval results showed that the R2 = 0.83 and RMSE = 0.18 compared to the field measured CWC, which confirmed the feasibility to alleviate the ill-posed inversion problem by the multiple priori information.

Monitoring the environmental status of mountainous or hilly areas is very important for their great influence on the global ecosystem and humanity. The enhanced vegetation index (EVI) has been widely used in environmental monitoring. It can reduce background and atmospheric noise via its feedback-based format. However, the application of EVI in mountainous areas will be limited, because EVI is greatly affected by topographic effects as its soil adjustment index is not in a band ratio format. To moderate the topographic effects on EVI, we modified the EVI by changing the soil adjustment index from a constant to a variable related to the incidence angle. In the evaluation of the modified EVI, three other well-known topographic correction methods, Sun-canopy-sensor (SCS), SCS with C-correction (SCS+C), and modified Minnaert (MM), were used for comparison. The results indicated that the modified EVI and SCS+C perform better than MM and SCS by visual comparison. Quantitatively, modified EVI, which has an effect similar to SCS+C in the low incidence angle regions, largely decreased the standard deviation of the same land features and the correlation between EVI and the cosine of the incidence angle. When the incidence angle exceeds 90 deg, SCS+C and other two topographic correction methods caused overcorrections. However, modified EVI solved this problem well due to its smaller increasing curvature than other three topographic correction methods. Moreover, compared to SCS+C, modified EVI better preserved the characters of land surface features.