Zhengyong Zhao

The soil respiration (Rs) of forests is a major component of global Rs, yet few studies have focused on it. This study aimed to estimate global forest Rs and its changes at a resolution of 30 m via an artificial neural network (ANN) model. Five input candidates representing forest type, climatic, soil, and geographical information, as well as 1472 satisfactory forest Rs records, were used to build the ANN model and evaluate the model performance via a 10‐fold cross‐validation scheme. Global forest change data sets were used to accurately define the extent of forests and their changes, which was achievable because of the dynamic information and high resolution (30 m) of the data sets. The results indicate that the average annual global forest Rs from 2000 to 2020, as estimated by the optimal ANN model with an r² value of 0.67 and a root‐mean‐square error of 252.6 g C m⁻² yr⁻¹, was 46.24 ± 5.86 Pg C yr⁻¹. From 2001 to 2019, the average theoretical annual global forest Rs loss was 0.22 ± 0.06 Pg C yr⁻¹ due to an average forest loss area of 23.4 million ha yr⁻¹. In addition, the annual Rs theoretically increased by 0.75 Pg C in 2012 due to a global forest gain area of 80.5 million from 2001 to 2012. The presented data sets of global forest Rs and its changes can provide an accurate benchmark for discussing the carbon cycle and climate change at global to regional scales, even when operating over a small forest area (i.e., dozens of ha), which is a scale that has been ignored in other global Rs studies.

Accurate soil organic carbon (SOC) maps are helpful for guiding forestry production and management. Different ecological landscape areas within a large region may have different soil–landscape relationships, so models specifically for these areas may capture these relationships more accurately than the global model for the entire study area. The aim of this study was to investigate the role of zonal modelling in predicting forest SOC and to produce highly accurate forest SOC distribution maps. The prediction objects were SOC at five soil depths (0–20, 20–40, 40–60, 60–80, and 80–100 cm). First, the forest type map and soil texture class map were used to divide the relative homogeneous regions in Shaoguan City, Guangdong Province, China. Second, seven terrain variables derived from a 12.5-m digital elevation model (DEM) and five vegetation variables generated from 10-m Sentinel-2 remote sensing images were used as predictors to develop regional artificial neural network (ANN) models for each homogeneous region, as well as a global ANN model for the entire study area (1000 sample points). Finally, 10-fold cross-validation was used to assess the ANN prediction model performance, and independent validation was used to evaluate the produced forest SOC prediction maps (194 additional samples). The cross-validation results showed that the accuracies of the regional models were better than that of the global model. Independent validation results also showed that the precision (R²) of 0- to 100-cm forest SOC maps generated by forest type modelling had an improvement of 0.05–0.15, and that by soil texture class modelling had an improvement of 0.07–0.13 compared to the map generated by the global model. In conclusion, delineating relatively homogeneous regions via simple methods can improve prediction accuracy when undertaking soil predictions over large areas, especially with complex forest landscapes. In addition, SOC in the study area is generally more abundant in broadleaf forest and clay areas, with overall levels decreasing with soil depth. Accurate SOC distribution information can provide references for fertilization and planting. Plants with particularly high soil fertility requirements may perhaps be planted in broadleaf forests or clay areas, and plants with particularly developed roots may require furrow application of a small amount of SOC.

The study on the spatial distribution of forest soil nutrients is important not only as a reference for understanding the factors affecting soil variability, but also for the rational use of soil resources and the establishment of a virtuous cycle of forest ecosystems. The rapid development of remote sensing satellites provides an excellent opportunity to improve the accuracy of forest soil prediction models. This study aimed to explore the utility of the Gaofen-1 (GF-1) satellite in the forest soil mapping model in Luoding City, Yunfu City, Guangdong Province, Southeast China. We used 1000 m resolution coarse-resolution soil map to represent the overall regional soil nutrient status, 12.5 m resolution terrain-hydrology variables to reflect the detailed spatial distribution of soil nutrients, and 8 m resolution remote sensing variables to reflect the surface vegetation status to build terrain-hydrology artificial neural network (ANN) models and full variable ANNs, respectively. The prediction objects were alkali-hydro-nitrogen (AN), available phosphorus (AP), available potassium (AK), and organic matter (OM) at five soil depths (0–20, 20–40, 40–60, 60–80, and 80–100 cm). The results showed that the full-variable ANN accuracy at five soil depths was better than the terrain-hydrology ANNs, indicating that remote sensing variables reflecting vegetation status can improve the prediction of forest soil nutrients. The remote sensing variables had different effectiveness for different soil nutrients and different depths. In upper soil layers (0–20 and 20–40 cm), remote sensing variables were more useful for AN, AP, and OM, and were between 10%–14% (R²), and less effective for AK at only 8% and 6% (R²). In deep soil layers (40–60, 60–80, and 80–100 cm), the improvement of all soil nutrient models was not significant, between 3 and 6% (R²). RMSE and ROA ± 5% also decreased with the depth of soil. Remote sensing ANNs (coarse resolution soil maps + remote sensing variables) further demonstrated that the predictive power of remote sensing data decreases with soil depth. Compared to terrain-hydrological variables, remote sensing variables perform better at 0–20 cm, but the predictive power decreased rapidly with depth. In conclusion, the results of the study showed that the integration of remote sensing with coarse-resolution soil maps and terrain-hydrology variables could strongly improve upper forest soil (0–40 cm) nutrients prediction and NDVI, green band, and forest types were the best remote sensing predictors. In addition, the study area is rich in AN and OM, while AP and AK are scarce. Therefore, to improve forest health, attention should be paid to monitoring and managing AN, AP, AK, and OM levels.

Stocks and stoichiometry of carbon (C), nitrogen (N), and phosphorus (P) in ultisols are not well documented for converted forests. In this study, Ultisols were sampled in 175 plots from one type of secondary forest and four plantations of Masson pine (Pinus massoniana Lamb.), Slash pine (Pinus elliottii Engelm.), Eucalypt (Eucalyptus obliqua L’Hér.), and Litchi (Litchi chinensis Sonn., 1782) in Yunfu, Guangdong province, South China. Five layers of soil were sampled with a distance of 20 cm between two adjacent layers up to a depth of 100 cm. We did not find interactive effects between forest type and soil layer depth on soil organic carbon (SOC), total nitrogen (TN), and total phosphorus (TP) concentrations and storages. Storage of SOC was not different between secondary forests and Eucalypt plantations, but SOC of these two forest types were lower than that in Litchi, Masson pine, and Slash pine plantations. Soil C:P was higher in Slash pine plantations than in secondary forests. Soil CNP showed a decreasing trend with the increase of soil depth. Soil TP did not show any significant difference among soil layers. Soil bulk density had a negative contribution to soil C and P stocks, and longitude and elevation were positive drivers for soil C, N, and P stocks. Overall, Litchi plantations are the only type of plantation that obtained enhanced C storage in 0–100 cm soils and diverse N concentrations among soil layers during the conversion from secondary forests to plantations over ultisols.

Cadmium (Cd) is a toxic metal and found in various soils, including forest soils. The great spatial heterogeneity in soil Cd makes it difficult to determine its distribution. Both traditional soil surveys and spatial modeling have been used to study the natural distribution of Cd. However, traditional methods are highly labor-intensive and expensive, while modeling is often encumbered by the need to select the proper predictors. In this study, based on intensive soil sampling (385 soil pits plus 64 verification soil pits) in subtropical forests in Yunfu, Guangdong, China, we examined the impacting factors and the possibility of combining existing soil information with digital elevation model (DEM)-derived variables to predict the Cd concentration at different soil depths along the landscape. A well-developed artificial neural network model (ANN), multi-variate analysis, and principal component analysis were used and compared using the same dataset. The results show that soil Cd concentration varied with soil depth and was affected by the top 0–20 cm soil properties, such as soil sand or clay content, and some DEM-related variables (e.g., slope and vertical slope position, varying with depth). The vertical variability in Cd content was found to be correlated with metal contents (e.g., Cu, Zn, Pb, Ni) and Cd contents in the layer immediately above. The selection of candidate predictors differed among different prediction models. The ANN models showed acceptable accuracy (around 30% of predictions have a relative error of less than 10%) and could be used to assess the large-scale Cd impact on environmental quality in the context of intensifying industrialization and climate change, particularly for ecosystem management in this region or other regions with similar conditions.

Soil organic carbon storage (SOCS) estimation is a crucial branch of the atmospheric–vegetation–soil carbon cycle study under the background of global climate change. SOCS research has increased worldwide. The objective of this study is to develop a two-stage approach with good extension capability to estimate SOCS. In the first stage, an artificial neural network (ANN) model is adopted to estimate SOCS based on 255 soil samples with five soil layers (20 cm increments to 100 cm) in Luoding, Guangdong Province, China. This method is compared with three common methods: The soil type method (STM), ordinary kriging (OK), and radial basis function (RBF) interpolation. In the second stage, a linear model is introduced to capture the regional differences and further improve the estimation accuracy of the Luoding-based ANN model when extending it to Xinxing, Guangdong Province. This is done after assessing the generalizability of the above four methods with 120 soil samples from Xinxing. The results for the first stage show that the ANN model has much better estimation accuracy than STM, OK, and RBF, with the average root mean square error (RMSE) of the five soil layers decreasing by 0.62–0.90 kg·m⁻², R² increasing from 0.54 to 0.65, and the mean absolute error decreasing from 0.32 to 0.42. Moreover, the spatial distribution maps produced by the ANN model are more accurate than those of other methods for describing the overall and local SOCS in detail. The results of the second stage indicate that STM, OK, and RBF have poor generalizability (R² < 0.1), and the R² value obtained with ANN method is also 43–56% lower for the five soil layers compared with the estimation accuracy achieved in Luoding. However, the R² of the linear models built with the 20% soil samples from Xinxing are 0.23–0.29 higher for the five soil layers. Thus, the ANN model is an effective method for accurately estimating SOCS on a regional scale with a small number of field samples. The linear model could easily extend the ANN model to outside areas where the ANN model was originally developed with a better level of accuracy.

Ecosites are required for stand-level forest management and can be determined within a two-dimensional edatopic grid with soil nutrient regimes (SNRs) and soil moisture regimes (SMRs) as coordinates. A new modeling method is introduced in this study to map high-resolution SNR and SMR and then to design ecosites in Nova Scotia, Canada. Using coarse-resolution soil maps and nine topo-hydrologic variables derived from high-resolution digital elevation model (DEM) data as model inputs, 511 artificial neural network (ANN) models were developed by a 10-fold cross-validation with 1507 field samples to estimate 10 m resolution SNR and SMR maps. The results showed that the optimal models for mapping SNR and SMR engaged eight and seven topo-hydrologic variables, together with three coarse-resolution soil maps, as model inputs, respectively; 82% of model-estimated SNRs were identical to field assessments, while this value was 61% for SMRs, and the produced ecosite maps had 67–68% correctness. According to the error matrix, the predicted SNR and SMR maps greatly alleviated poor prediction in the areas of extreme nutrient or moisture conditions (e.g., very poor or very rich, wet, or very dry). Thus, the new method for modeling high-resolution SNR and SMR could be used to produce ecosite maps in sites where accessibility is hard.

High-resolution maps of soil property are considered as the most important inputs for decision support and policy-making in agriculture, forestry, flood control, and environmental protection. Commonly, soil properties are mainly obtained from field surveys. Field soil surveys are generally time-consuming and expensive, with a limitation of application throughout a large area. As such, high-resolution soil property maps are only available for small areas, very often, being obtained for research purposes. In the chapter, artificial neural network (ANN) models were introduced to produce high-resolution maps of soil property. It was found that ANNs can be used to predict high-resolution soil texture, soil drainage classes, and soil organic content across landscape with reasonable accuracy and low cost. Expanding applications of the ANNs were also presented.

Forest ecosite reflects the local site conditions that are meaningful to forest productivity as well as basic ecological functions. Field assessments of vegetation and soil types are often used to identify forest ecosites. However, the production of high-resolution ecosite maps for large areas from interpolating field data is difficult because of high spatial variation and associated costs and time requirements. Indices of soil moisture and nutrient regimes (i.e., SMR and SNR) introduced in this study reflect the combined effects of biogeochemical and topographic factors on forest growth. The objective of this research is to present a method for creating high-resolution forest ecosite maps based on computer-generated predictions of SMR and SNR for an area in Atlantic Canada covering about 4.3 × 10⁶ hectares (ha) of forestland. Field data from 1,507 forest ecosystem classification plots were used to assess the accuracy of the ecosite maps produced. Using model predictions of SMR and SNR alone, ecosite maps were 61 and 59% correct in identifying 10 Acadian- and Maritime-Boreal-region ecosite types, respectively. This method provides an operational framework for the production of high-resolution maps of forest ecosites over large areas without the need for data from expensive, supplementary field surveys.

Biome‐specific soil respiration (Rs) has important yet different roles in both the carbon cycle and climate change from regional to global scales. To date, no comparable studies related to global biome‐specific Rs have been conducted applying comprehensive global Rs databases. The goal of this study was to develop artificial neural network (ANN) models capable of spatially estimating global Rs and to evaluate the effects of interannual climate variations on 10 major biomes. We used 1976 annual Rs field records extracted from global Rs literature to train and test the ANN models. We determined that the best ANN model for predicting biome‐specific global annual Rs was the one that applied mean annual temperature (MAT), mean annual precipitation (MAP), and biome type as inputs (r² = 0.60). The ANN models reported an average global Rs of 93.3 ± 6.1 Pg C yr⁻¹ from 1960 to 2012 and an increasing trend in average global annual Rs of 0.04 Pg C yr⁻¹. Estimated annual Rs increased with increases in MAT and MAP in cropland, boreal forest, grassland, shrubland, and wetland biomes. Additionally, estimated annual Rs decreased with increases in MAT and increased with increases in MAP in desert and tundra biomes, and only significantly decreased with increases in MAT (r² = 0.87) in the savannah biome. The developed biome‐specific global Rs database for global land and soil carbon models will aid in understanding the mechanisms underlying variations in soil carbon dynamics and in quantifying uncertainty in the global soil carbon cycle.