Forest Method Research Articles

Predicting overall survival (OS) in patients (pts) with penile squamous cell carcinoma (PSCC) undergoing regional lymph node dissection (LND) ± multimodal therapy.

512 Background: Advances in prognostic allocation of pts with PSCC receiving regional LND are needed, along with an assessment of the role of perioperative treatments. Methods: An international, multicenter, retrospective database was queried for all pts with PSCC who had received inguinal ± pelvic LND from 1980 to 2017. We classified or re-classified pts according to the 2009 TNM classification system. The primary endpoint was OS. Survival analysis methods were adopted for the purpose of detecting and modeling putative prognostic factors, and develop a prognostic stratification tool. The analysis was performed in two steps: first, we modeled outcome data and covariates by resorting to a random forest (RF) method (an “ensemble” machine-learning approach, with an ad hoc approach incorporating missing data imputation). Second, data was modeled using Cox proportional hazard regression. In addition, a Cox model was fit including pre-specified variables that were deemed to be clinically relevant. Results: There were 743 pts who received LND at seven referral centers from Europe, the USA, Brazil, and China. Of these pts, 689 were analyzed: 86 (12.5%) received neoadjuvant chemotherapy (NAC); 171 (24.8%) received adjuvant chemotherapy (AC), and 74 (10.7%) received adjuvant radiotherapy. The variables significantly associated with OS were age (p < 0.001), the pathologically involved/total removed LN ratio (p < 0.001), pN stage (overall p < 0.001) and NAC (HR: 1.58, 95%CI: 1.10-2.26, p = 0.013). NAC and AC were ineffective or detrimental in N1-2 patients, whereas they provided OS improvements in N3 patients. OS improvement reached statistical significance for AC in pelvic LN-positive pts (p = 0.046). Finally, we developed a nomogram predicting 12- and 24-month OS based on pre-specified variables (c-index: 0.75). The study is limited by its retrospective nature. Conclusions: We propose a post-treatment prognostic tool that can be offered as an aid to physicians to enhance decision-making, clinical research, and patient counseling after LND in PSCC. Administration of NAC and AC should be restricted to N3 patients.

Automated extraction and discrimination of open land areas from IRS-1C LISS III imagery

Availability of open land area is a key indicator in assessing and planning urban environments. However, accurate uninhabited surface extraction is still a challenge. In this paper an attempt is made to combine the advantages of partial differential equation (PDE) and random forest (RF) method for segmentation of IRS-1C LISS III satellite image for mapping and discrimination of open land areas. Spatial variations of the pixels in the image are analyzed to identify open land pixels. PDE method denoise and conserve finer details simultaneously by using correlation between the spectral bands and directions of the edges. Low-resolution resultant image patch is mapped to high-resolution image patch by linear regression model. Variance of classification is reduced by training many classifiers using interpolated RF method. This methods elevates the accuracy of the direct RF method and achieves 3.35 dB improvement in PSNR and 6.47 reduction in MSE. Further, discrimination of open land areas is done into distinct classes, using inherent spatial information. Accuracy assessment indicates an overall accuracy of 87% over direct RF method.

Identifikasi Pendangkalan Sungai Kahayan Akibat Sedimen

The population growth rate continues to increase directly proportional to the rate of environmental damage. This is due to the increasing need for clothing, food, and shelter. Or primary needs and secondary needs. For example, the impact of floating houses on pollution of the Kahayan River in the City of Palangka Raya, Central Kalimantan. One example of the impact that will be caused is river siltation, river water pollution, and the generation of garbage around the river. The purpose of this study was to identify the silting of the Kahayan River due to sedimentation and to determine the crossing of the Kahayan River in the study area. The research location was in the Kahayan Palangkaraya Tugu Soekarno River Catchment Area to Ropi Ropi. In the sedimentation calculation using USLE modification (MUSLE) has calculated the results of the increase in sediment in the river. The Kahayan River has an area of 81,648 km2, a length of 600 km, a width of 500m, its depth reaches 7 m and empties into three regencies between the cities of Palangkaraya, Kabupaten Gunung Mas, and Kabupaten Pulang Pisau. From the calculation of flood discharge, the peak discharge (Qp) is 0.0105 km�. Prediction of Sedimentation Rate using the MUSLE Method of each Protection Forest = 0, Sebangau National Park = 901, Tourist Park = 73, Forest Park = 112, Tetat Production Forest = 3.422

SAR and QSAR models of cyclooxygenase-1 (COX-1) inhibitors

ABSTRACTCyclooxygenase-1 (COX-1) is one isoform of COX, and it is a main target of nonsteroidal anti-inflammatory drugs (NSAIDs). It is important to develop efficient and selective COX-1 inhibitors. In this work, 12 classification models for 1530 cyclooxygenase-1 (COX-1) inhibitors were built by support vector machine (SVM), decision tree (DT) and random forest (RF) methods. The best classification model (model 1A) was built by SVM with MACCS fingerprints. The classification accuracies for the training and test sets were 99.67% and 97.39%, respectively. The Matthews correlation coefficient (MCC) of the test set was 0.94. We also divided the 1530 COX-1 inhibitors into nine subsets according to their different scaffolds using Kohonen’s self-organizing map (SOM). In addition, six quantitative structure–activity relationship (QSAR) models for 181 COX-1 inhibitors whose IC50 were measured by enzyme immunoassay were built by multiple linear regression (MLR) and SVM. The best QSAR model (model 5A) was built by SVM with CORINA Symphony descriptors. The correlation coefficients of the training and test sets are 0.93 and 0.84, respectively. The models built in this study can be obtained from the authors.

Historical cost vs fair value in forest accounting: the case of Lithuania

The accounting methods based on two concepts-accounting at historical cost and accounting at fair value-are responsible for the provision of information about the enterprise's assets in the financial statements. Taking into consideration the specifics of the national financial accounting regulation, the forests managed by the forestry enterprises can also be measured by using either of these two accounting methods. However, in terms of accounting, both of them pose certain problems or ambiguity. The purpose of the research is to evaluate the strengths and weaknesses of forest accounting methods based on fair value and historical cost and the practice of their use in Lithuanian forestry enterprises. The study examines scholarly literature and deploys the theoretical methods of comparative analysis, critical evaluation, systematisation, generalisation. The empirical research involved document content analysis, questionnaire survey. The article deals with the issues of the use of accounting methods for forestry accounting: traditional cost-based accounting methods do not reflect the biological forest transformation, hinders identifying the forest development costs and the end of their capitalisation, the method of a systematic derecognition. On the other hand, the essential complication of the use of the fair value method is that the forest largely lacks an active market with quoted prices. Thus, its fair value is determined on the basis of rather subjective assumptions by means of diverse valuation methods, resulting in unreliable and unverifiable information. The results of the research carried out into the forestry accounting policy observed in the Lithuania's private forestry enterprises revealed that forest accounting by cost is exclusively carried out by all the enterprises under investigation. Nevertheless, the method itself was interpreted quite differently. The article presents the modified forestry accounting methods by cost, which allow reducing the identified shortcomings.

Comparison of ALS- and UAV(SfM)-derived high-density point clouds for individual tree detection in Eucalyptus plantations

ABSTRACTHighly accurate, rapid forest inventory techniques are needed to enable forest managers to address the increasing demand for sustainable forestry. In the last two decades, Airborne Laser Scanning (ALS) and Terrestrial Laser Scanning have become internationally established as forest mapping and monitoring methods. However, recent advances in sensors and in image processing – particularly Structure from Motion (SfM) technology – have also enabled the extraction of dense point clouds from images obtained by Digital Aerial Photography (DAP). DAP is cheaper than ALS, especially when the systems are mounted on small unmanned aerial vehicles (UAVs), and the density of the point cloud can easily reach the levels yielded by ALS devices. The main objective of this study was to evaluate and compare the usefulness of ALS-derived and UAV(SfM)-derived high-density point clouds for detecting and measuring individual tree height in Eucalyptus spp. plantations established on complex terrain. A total of 325 reference trees were measured and located in 6 square plots (400 m2). The individual tree crown (ITC) delineation algorithm detected 311 from the ALS-derived data and 259 trees from the UAV(SfM)-derived data, representing accuracy levels of, respectively, 96% and 80%. The results suggest that at plot level, UAV(SfM)-generated point clouds are as good as ALS-derived point clouds for estimating individual tree height. Furthermore, analysis of the differences in digital elevation models at landscape level showed that the elevations of the UAV(SfM)-derived terrain surfaces were slightly higher than the ALS-derived surfaces (mean difference, 1.14 m and standard deviation, 1.93 m). Finally, we discuss how non-optimal UAV-image-acquisition conditions and slope terrain affect the ITC delineation process.

LAS CODORNICES DE BOSQUES Y SELVAS. RETOS DE MANEJO Y CONSERVACIÓN DE LA FAUNA SILVESTRE


 
 
 Las codornices que habitan en bosques y selvas representan un buen modelo para analizar los retos que representan tanto el manejo como la conservación de la fauna silvestre, debido a su estrecha dependencia hacia esos ecosistemas. Las políticas gubernamentales de uso y protección de los recursos naturales repercuten en las condiciones en las que se mantiene la vegetación como hábitat, por lo que aquí se examinan esos factores y su posible influencia en estas aves, como un caso de estudio. De especial importancia son las Unidades de Manejo para la Conservación de la Vida Silvestre (UMA), por ser un avance respecto a los anteriores enfoques puramente cinegéticos, ya que pretenden promover la revaloración de los recursos naturales, la diversificación productiva de los poseedores de la tierra y su participación en ellas. Sin embargo, el modelo sigue siendo polémico y aún tiene serios problemas de implementación, por la falta de rigor científico en el monitoreo de las poblaciones que se intenta aprovechar. En general, el principal reto consiste en mantener el hábitat mediante la conservación de la vegetación arbolada. Esto implica la inclusión de prácticas sustentables en la aplicación de métodos de manejo forestal que consideren las necesidades de la fauna, así como la disminución de las tasas de deforestación. Las codornices forestales justifican un mayor interés y atención, ya que la comprensión de su biología y ecología es muy limitada.
 
 

Progressive Amalgamation of Building Clusters for Map Generalization Based on Scaling Subgroups

Map generalization utilizes transformation operations to derive smaller-scale maps from larger-scale maps, and is a key procedure for the modelling and understanding of geographic space. Studies to date have largely applied a fixed tolerance to aggregate clustered buildings into a single object, resulting in the loss of details that meet cartographic constraints and may be of importance for users. This study aims to develop a method that amalgamates clustered buildings gradually without significant modification of geometry, while preserving the map details as much as possible under cartographic constraints. The amalgamation process consists of three key steps. First, individual buildings are grouped into distinct clusters by using the graph-based spatial clustering application with random forest (GSCARF) method. Second, building clusters are decomposed into scaling subgroups according to homogeneity with regard to the mean distance of subgroups. Thus, hierarchies of building clusters can be derived based on scaling subgroups. Finally, an amalgamation operation is progressively performed from the bottom-level subgroups to the top-level subgroups using the maximum distance of each subgroup as the amalgamating tolerance instead of using a fixed tolerance. As a consequence of this step, generalized intermediate scaling results are available, which can form the multi-scale representation of buildings. The experimental results show that the proposed method can generate amalgams with correct details, statistical area balance and orthogonal shape while satisfying cartographic constraints (e.g., minimum distance and minimum area).

Multi-sensor forest vegetation height mapping methods for Tanzania

ABSTRACTThis paper proposes a new method for mapping of forest cover in Tanzania, in the form of yearly estimates of average vegetation height from time-series of Landsat and ALOS PALSAR satellite images. By using airborne laser scanning data and Landsat-8 data from 2014, a regression between average vegetation height and the specific leaf area vegetation index is established. By using all available Landsat acquisitions of the same area within 1 year, and producing a yearly estimate of vegetation height, the estimation error variance is reduced. The variance is further reduced by Kalman filtering the sequence of yearly estimates. A multi-sensor version of the method comprises application of the radar backscatter when L-band SAR data is available. To conclude, we have demonstrated that estimation of mean vegetation height is possible from dense time series of optical and SAR satellite data. Change detection was able to detect areas with total loss of biomass.

Prediction of diabetic retinopathy based on a committee of random forests

Diabetic retinopathy (DR) is an ocular disease generated by complications of diabetes, and it must be discovered quickly for effective cure. By early diagnosis of retinal fundus disease, ophthalmologists can cure the disease or reduce its deterioration, thereby preventing the patients from vision loss. Using enlarged images, ophthalmologists can diagnose DR. In this paper, committee of random forests (CRF) for detection of DR is proposed. In this approach, we use k-means clustering algorithm and random forest classification method to create a new classifier. CRF has been tested on the DR Debrecen dataset, in which 94.76% accuracy is reached in disease or no disease setting.

GROWTH AND PRODUCTIVITY OF HYBRID POPLAR PLANTANIONS IN CONDITIONS OF WESTERN FOREST-STEPPE OF UKRAINE

Досліджено особливості росту гібридних тополь в умовах свіжого груду Західного Лісостепу України до 9-річного віку з огляду на доцільність їх використання у короткоротаційних плантаціях. Насадження створено однорічними живцями клонів: 'Дружба' 'Канадська × Бальзамічна' 'Стрілоподібна' і 'Тронко'. Ґрунт – сірий лісовий свіжий. Розміщення садивних місць – 2,0×0,8 м. Під час проведення досліджень застосовано традиційні лісівницькі та таксаційні методи. Найвищу збереженість рослин у 7-річному віці виявлено у клонів 'Тронко' (88 %) і 'Дружба' (86 %), найменшу – у клону 'Канадська × Бальзамічна' – 78 %. Найбільшу середню висоту мав культивар 'Дружба' (10,3±0,33 м). Решта клонів за висотою практично не відрізнялися. Найбільший середній діаметр дерев (7,4±0,33 см) встановлено у клону 'Канадська х Бальзамічна', який мав найменшу збереженість. Найвищий запас визначено у деревостанів клонів 'Канадська × Бальзамічна' (113 м3∙га-1)і 'Дружба' (122 м3∙га-1). Отже, 7-річні насадження продукували від 13,3 до 17,4 м3∙га-1деревини на рік. За наступні 2 роки відбувся значний відпад дерев клонів 'Тронко' і 'Канадська × Бальзамічна', що негативно позначилось на показниках їх запасу. Найвищим запас був у клону 'Дружба' – 139 м3∙га-1. Середня зміна запасу та загальної продуктивності (запас + відпад) усіх клонів виявилися меншими у віці 9 років, ніж у 7, що свідчить про те, що в досліджуваних умовах оптимальний вік рубання деревостанів тополі густотою близько 6 тис. шт. на 1 га становить 6-7 років. При цьому, насадження клону 'Дружба' мають запас 122, 'Канадська × Бальзамічна' – 113, 'Тронко' – 105,'Стрілоподібна' – 93 м3∙га-1.

Abstract 21081: Coffee Intake Affects Heart Failure and Stroke Survival and is Significant in Predicting Heart Failure and Stroke Risk

Introduction: Heart Failure (HF) and stroke are complex diseases with multiple phenotypes. While many risk factors for these diseases are well known, it is likely that there are as-yet unidentified risk factors given the complex pathophysiology of each disease. We investigated the diet domain in the Framingham Heart Study (FHS), Cardiovascular Heart Study (CHS), and Atherosclerosis Risk In Communities (ARIC) study to identify potential diet factors associated with HF and stroke. Methods: We used random forest machine-learning methods to identify risk factors associated with cardiovascular diseases in the Framingham Heart Study. We used Cox proportional hazards analysis to investigate these findings further in the FHS and validated our findings using the CHS and ARIC datasets. Results: Random forest analysis identified known risk factors such as blood pressure, age, and cholesterol. In addition, coffee and red meat intake were also found to be important in predicting heart failure and stroke. Survival analysis confirmed these results suggesting coffee intake is significantly associated with HF and stroke. Coffee decreased risk with HF and stroke by 7% and 8% respectively with every increased cup consumed per week (p = .001 for both HF and Stroke). Validation of this trend was confirmed in the CHS and ARIC which resulted in similar outcomes. Furthermore, we built a risk model using known risks from the Framingham Risk Score guidelines and predicted CHF and stroke outcomes. By including coffee in the model, the prediction accuracy increased by 4%. Conclusion: Increased coffee intake is associated with reduced HF and stroke outcomes in FHS, CHS and ARIC.

Assessment of crown fire initiation and spread models in Mediterranean conifer forests by using data from field and laboratory experiments

Aims of study: To conduct the first full-scale crown fire experiment carried out in a Mediterranean conifer stand in Spain; to use different data sources to assess crown fire initiation and spread models, and to evaluate the role of convection in crown fire initiation.Area of study: The Sierra Morena mountains (Coordinates ETRS89 30N: X: 284793-285038; Y: 4218650-4218766), southern Spain, and the outdoor facilities of the Lourizán Forest Research Centre, northwestern Spain.Material and methods: The full-scale crown fire experiment was conducted in a young Pinus pinea stand. Field data were compared with data predicted using the most used crown fire spread models. A small-scale experiment was developed with Pinus pinaster trees to evaluate the role of convection in crown fire initiation. Mass loss calorimeter tests were conducted with P. pinea needles to estimate residence time of the flame, which was used to validate the crown fire spread model.Main results: The commonly used crown fire models underestimated the crown fire spread rate observed in the full-scale experiment, but the proposed new integrated approach yielded better fits. Without wind-forced convection, tree crowns did not ignite until flames from an intense surface fire contacted tree foliage. Bench-scale tests based on radiation heat flux therefore offer a limited insight to full-scale phenomena.Research highlights: Existing crown fire behaviour models may underestimate the rate of spread of crown fires in many Mediterranean ecosystems. New bench-scale methods based on flame buoyancy and more crown field experiments allowing detailed measurements of fire behaviour are needed. p>ERRATUM PDF

Path to Create the Democracy Ways of Self-Governance in the Ethnic Communities in Thailand's North-Eastern Region

The study aimed to take lessons and experience from the community forest management methods of three ethnic communities namely, Phu Tai, Tai Lao and Tai Kui, and to study the guidelines of forest resource management among ethnic communities in the Northeastern region of Thailand, based on the on self-management democracy in communities. The researcher used qualitative research methodology in the three major ethnic communities in the Northeast, namely: 1) the Phu Tai ethnic community residing in the community forest, Nong Hang, Kalasin; 2) the Tai Lao ethnic communities residing in Phu Kham Noi Community Forest, Sawawat, Khon Kaen, and 3) the Tai Kui ethnic communities residing in the Don Pu Ta Baan Arlue Forest Village, Sum Rong Taab, Surin. The researcher conducted the documentation analysis and field research by participatory observation and in-depth interviewing the key informants who are involving in community forest management in each area amount total 30 peoples. All the data were analyzed by using the content analysis model and compare the case studies’ information for providing to the conclusion of the descriptive research results. The research findings were as follows: 1) based on the lessons learned and experiences from community forest management methods used as case studies, the self-management in democratic way can be summarized into four patterns as follows: 1) protective democratic, 2) direct democracy, 3) participatory democracy, and 4) discussed democracy. However, there is no single model used to analyze self-management in democratic way of ethnic communities as the case studies of this research. There are the needs of the development of participatory democracy with the integrated democratic practices. As well as the management of community forest in the selected ethnic communities can be characterized by values and features of democratic practices into seven forms. These are the unique features of the hidden democracy in the community forest management process. In addition, community forest management is developed as a part of learning process of democratic practices as the mentioned seven principles based on community culture. Lastly, the management of forest resources among the ethnic communities in the Northeastern region are based on self-management with democratic principles in line with the two practices. These mechanisms lead to the concept of responsibility, citizenship, and the creation of community-based roles in terms of self-management. This will lead to the establishment of a democratic pathway that is appropriate to the local community, which contributes to the stability and security of the national democratic system.

Supervised SVM Classification of Rainfall Datasets

Objectives: The model built in this paper is used to classify the rainfall datasets in identifying districts of more rainfall and of lesser rainfall in the state of Andhra Pradesh. Methods: In this paper support vector machine, random forest, Knearest neighbor and decision tree classification methods have been used to classify rainfall data sets which is divided into training set and test set for classification and later validation of the obtained results. Findings: Based on various statistical parameters like sensitivity, prevalence, detection rate, specificity, and detection prevalence it has been concluded that support vector machine classification methods is better than any other classification method used in the research. Rainfall data sets are used to initially build the classification model and the results are tested against the test set. Using the confusion matrix thus obtained the mentioned statistical parameters are obtained to establish the supremacy of support vector machine classification method. Applications: Examples of satellite imagery has become ever more significant in numerous application domains such as ecology monitoring and alternative discovery. Rainfall classification is the application used herein.

Grain size distribution for identifying the depositional microfacies of sediments in Luan River catchment

This study was performed in order to establish new parameters to recognize the depositional microfacies using grain size distribution (GSD) data from laser diffraction analysis. A total of 938 samples were collected from five sedimentary microfacies, and 18 characteristic parameters of these samples were developed by statistical analysis, fractal theory, fingerprint methods, and variogram function. The samples were divided into a training data set and testing data set. The training data set was used to establish a discriminative database with five depositional microfacies; the testing data set was used to verify whether these characteristic parameters can identify their depositional microfacies according to the discriminative database. Out of the 50 samples in the testing data set, 47 samples were correctly recognized within their depositional environment and only three samples of point bar were shared with natural levee using the random forest (RF) method. The analysis results indicate the fine grain size and general characteristic of GSD have more significant influences on depositional environment classification than other parameters. The flood plain and fan-between depression environments had the most similar hydrodynamic conditions among the five considered depositional microfacies. Therefore, the 18 characteristic parameters of GSD are useful tools to identify depositional microfacies for hydrogeological parameter studies.

SAR and QSAR study on the bioactivities of human epidermal growth factor receptor-2 (HER2) inhibitors

In this paper, structure–activity relationship (SAR, classification) and quantitative structure–activity relationship (QSAR) models have been established to predict the bioactivity of human epidermal growth factor receptor-2 (HER2) inhibitors. For the SAR study, we established six SAR (or classification) models to distinguish highly and weakly active HER2 inhibitors. The dataset contained 868 HER2 inhibitors, which was split into a training set including 580 inhibitors and a test set including 288 inhibitors by a Kohonen’s self-organizing map (SOM), or a random method. The SAR models were performed using support vector machine (SVM), random forest (RF) and multilayer perceptron (MLP) methods. Among the six models, SVM models obtained superior results compared with other models. The prediction accuracy of the best model (model 1A) was 90.27% and the Matthews correlation coefficient (MCC) was 0.80 on the test set. For the QSAR study, we chose 286 HER2 inhibitors to establish six quantitative prediction models using MLR, SVM and MLP methods. The correlation coefficient (r) of the best model (model 4B) was 0.92 on the test set. The descriptors analysis showed that HAccN, lone pair electronegativity and π electronegativity were closely related to the bioactivity of HER2 inhibitors.

Multi-model estimation of understorey shrub, herb and moss cover in temperate forest stands by laser scanner data

In temperate forests, the highest plant richness is regularly found in the understorey, i.e. shrub, tree regeneration, herbal and moss covers, which provides important food and shelter for other plant and animal species. Here, Light Detection And Ranging (LiDAR) remote sensing was investigated as a surrogate to laborious field surveys to improve understanding of the causal and predictive attributes of understorey.We designed a study in which we used a high-density LiDAR point cloud and applied a thinning algorithm to simulate two lower density point clouds including first and last returns and half of the remaining points (half-thinned data) and only first and last returns (F/L-thinned data). From each dataset, several over- and understorey-related statistical metrics were derived. Each of the three sets of LiDAR metrics was then combined with the forest habitat information to estimate the recorded proportions of shrub, herb and moss coverages.We used three different model procedures including zero-and-one-inflated beta regression (ZOINBR), ordinary least squares with logittransformed response variables (logistic model) and a machine learning random forest (RF) method. The logistic and ZOINBR model results showed highly significant relationships between LiDAR metrics and habitat types in explaining understorey coverage. The highest coefficients of determination included r2 = 0.80 for shrub cover (estimated by F/L-thinned data and ZOINBR model), r2 = 0.53 for herb cover (estimated by half-thinned data and logistic model) and r2 = 0.48 for moss cover (estimated by half-thinned data and logistic model). RF models returned the best predictive performances (i.e. the lowest root mean square errors). Despite slight differences, no substantial difference was observed amongst the performances achieved by the original, halfthinned and F/L-thinned point clouds. Moreover, the ZOINBR models did not improve predictive performances compared with the logistic model, which suggests that the latter should be preferred due to its greater simplicity and parsimony. Despite the differences between our simulated data and the real-world LiDAR point clouds of different point densities, the results of this study are thought to mostly reflect how LiDAR and forest habitat data can be combined for deriving ecologically relevant information on temperate forest understorey vegetation layers. This, in turn, increases the applicability of prediction results for overarching aims such as forest and wildlife management.

Importance of woody materials for seasonal variation in leaf area index from optical methods in a deciduous needleleaf forest

ABSTRACTWoody materials (woody area index, WAI) is a key error source in estimating leaf area index (LAI) by optical methods, but how to correct the error caused by WAI during different seasons has not reached consensus. In this study, effective plant area index (PAIe) was first estimated using two indirect optical methods (digital hemispherical photography, DHP, and LAI-2000) in a deciduous needleleaf forest, and then four different schemes for correcting the contribution of WAI to PAIe were tested here. We also directly estimated the seasonality of LAI by a litter collection method and an allometric method. Directly subtracting WAI from PAI resulted in a greater degree of uncertainty in correcting seasonal changes of PAIe from both DHP and LAI-2000. Therefore, we introduced a new correction factor, the stem-to-total area ratio, which was reasonable and useful for quantifying seasonal changes in the contribution of WAI to PAIe. We finally recommend a practical scheme for correcting PAIe from both DHP and LAI-2000, with accuracies as high as 88% and 87% during most growing seasons, respectively. Additionally, LAI values estimated from allometry were concordant with those estimated from litter collection, indicating that the allometry method is useful for tracking seasonal changes in LAI.

Tree-based techniques to predict soil units

Quantitative soil–landscape models offer a method for conducting soil surveys that use statistical tools to predict natural patterns in the occurrence of particular map units across a landscape. The aim of the present study was to predict soil units in a watershed with wide variation in landscape conditions. The approach relied on a modelling of soil-forming factors in order to understand the variability of the landscape components in the region. Models were generated for landscape attributes related to pedogenesis, specifically elevation, slope, curvature, compound topographic index, Euclidean distance from stream networks, landforms map, clay minerals index, iron oxide index and normalised difference vegetation index, along with an existing geology map. The soil classification was adapted from the World Reference Base System for Soil Resources, and the predominant soil taxonomic orders observed were Ferrasols, Acrisols, Gleysols, Cambisols, Fluvisols and Regosols. The algorithms used to predict the soil units were based on decision tree (DT) and random forest (RF) methods. The criteria used to evaluate the models’ performance were statistical indices, coherence between predicted units and the legacy map, as well as accuracy checks based on control samples. The best performing model was found to be the RF algorithm, with resulting statistical indices considered excellent (overall=0.966, kappa=0.962). The accuracy of the map as determined by control points was 67.89%, with a kappa value of 61.39%.