Statistical modelling and forecasting annual sugarcane production in India: Using various time series models

This paper proposes a comparison of various time series forecasting models to forecast annual data on sugarcane production over 63 years from 1960 to 2022. In this research, the Mean Forecast Model, the Naive Model, the Simple Exponential Smoothing Model, Holt's model, and the Autoregressive Integrated Moving Average time series models have all been used to make effective and accurate predictions for sugarcane. Different scale-dependent error forecasting techniques and residual analysis have been used to examine the forecasting accuracy of these time series models. SE of Residuals, Root Mean Squared Error (RMSE), Mean Absolute Error (MAE) and Akaike's Information Criterion (AIC) are used to analyse the forecast's accuracy. The best model has been selected based on the predictions with the lowest value, according to the three-performance metrics of RMSE, MAE, and AIC. The estimated sugarcane production shows an increasing trend for the next 10 years and is projected to be 37,763.38 million tonnes in the year 2032. Further, empirical results support the plan and execution of viable strategies to advance sugarcane production in India to fulfil the utilisation need of the increasing population and further improve food security.     

木聚糖酶氨基酸组成与其最适pH的神经网络模型

籍均匀设计（UD）方法,构建了G/11家族木聚糖酶氨基酸组成和最适pH的神经网络（NNs）模型。当学习速率为0.09、动态参数为0.4、Sigmoid参数为0.98,隐含层结点数为10时,该模型对最适pH的拟合和预测平均绝对百分比误差可分别达到3.02%和4.06%,均方根误差均为0.19个pH单位,平均绝对误差分别为0.11和0.19个pH单位。该结果比文献报道的用逐步回归方法好。     

基于均匀设计的主成分分析-支持向量机模型及其在几丁质酶最适pH建模中的应用

采用主成分分析法对样本数据集进行预处理,将得到的新样本数据集输入支持向量机,籍均匀设计,构建了几丁质酶氨基酸组成和最适pH的数学模型。当惩罚系数C为10,epsilon值为0.7,Gamma值为0.5,模型对pH值拟合的平均绝对百分比误差为3.76%,同时具有良好的预测效果,预测的平均绝对误差为0.42个pH单位。该方法比用BP神经网络方法效果更佳。     

基于辅助环境变量的土壤有机碳空间插值——以黄土丘陵区小流域为例

利用普通克里格法(OK)、反距离加权法(IDW)、径向基函数法(RBF)、基于土地利用类型修正的普通克里格法(OK_LU)4种插值方法,对黄土丘陵羊圈沟小流域的土壤有机碳含量进行空间插值。预测结果的准确性通过Pearson相关系数(R),平均绝对误差(MAE),均方根误差(RMSE),准确度(AC)来评价。研究结果表明:(1)在前3种常规空间插值方法中,OK对刻画区域土壤有机碳的空间分布趋势效果最佳,其预测MAE值和RMSE值均为最小,Pearson相关系数(R)和准确度(AC)最大,说明其预测结果的准确性最好、预测的极端误差也最小;其次为RBF;IDW预测的效果最差。(2)OK_LU在空间特征表达方面能够更好地反映复杂地形区的局部变异,其插值结果的精度相比OK有一定程度的提高,其平均绝对误差(MAE)从0.900%降到了0.567%,均方根误差(RMSE)从1.101%降到了0.777%,Pearson相关系数(R)从0.4026提高到0.5589,准确度(AC)从0.9081提高到0.9505。综合比较,在黄土丘陵地区,OK_LU能使插值结果的精度有较大提高,是土壤有机碳空间制图的有效途径。     

西藏自治区草地地上生物量遥感反演研究

草地地上生物量（Aboveground Biomass,AGB）是反映草地生态系统功能和质量的关键指标,大尺度地准确估算草地AGB对草地生态系统经营管理至关重要。研究以MODIS影像为数据源,提取反射率、植被指数和植被产品三种不同类型的特征变量,结合野外实测样地草地AGB数据,构建以多元线性逐步回归为代表的参数模型以及随机森林、支持向量机和kNN等非参数模型进行西藏自治区草地AGB估测及空间分布制图。结果表明：（1）多元线性逐步回归、随机森林、支持向量机和kNN模型在加入植被产品特征变量后,RMSE分别降低了15.8%、13.5%、4.1%和17.3%,表明植被产品作为建模变量用于草地AGB估测可有效提高模型精度;（2）三种组合变量构建的草地AGB估测模型中,反射率、植被指数、植被产品组合构建的模型效果最佳,其中kNN模型估测精度最高,R²达到0.60,RMSE和MAE分别为0.43、0.34 t/hm²;（3）草地AGB空间分布呈现出西北地区较低、中部地区较高且分布形态较破碎和东部地区较高的变化特征;（4）利用MODIS植被产品结合kNN模型的预测值与草地实测的AGB空间分布趋势基本一致。综上,MODIS植被产品结合kNN模型可作为大尺度区域草地AGB遥感估测的有效参考。     

Area-based scenario development in land-use change modeling: A system dynamics-assisted approach for mixed agricultural-residential landscapes

This study aimed to enhance land use and land cover (LULC) change models by addressing their main limitations, which include the lack of accountability and temporal stability of driving forces. Additionally, the study aimed to create area-based scenarios to forecast future LULCs, rather than solely relying on distribution-based scenarios. To accomplish this goal, the study developed a coupled System Dynamics (SD) and Cellular Automata (CA) modeling system to simulate possible LULC changes in the Gavkhooni Basin, central Iran. The study utilized LULC maps from Landsat images in 2001, 2011, and 2021 to analyze spatio-temporal land use changes in the region. Agricultural and residential transition suitability layers were produced using a spatial Multi-Criteria Evaluation procedure and applied to inform the CA model in the proper allocation of LULC changes. Three interconnected water supply, agricultural, and residential area projection subsystems were developed using system dynamics method to determine land requirements for LULC conversions from 2020 to 2041, taking into account factors such as water availability, land suitability, agricultural labor force, and economic development. Ten scenarios were developed based on changes in the key variables affecting the limiting factors, such as climatic conditions and water management policies, to project agricultural and residential areas in the future. The CA's spatial allocation informed by transition suitability layers was found to be satisfactory with a Kappa-location value of 0.85. The subsystems were competent in projecting water supply with Mean Absolute Error (MAE) values of 6.57% and the dynamics of agricultural and residential areas with MAE values of 2.94%, whereas those of the Markovian Chain model were found to be 23.02% and 7.5% for agricultural and residential areas, respectively. The study found that available agricultural areas varied significantly between 86.53 and 1480 sq.km under different climatic conditions, irrigation efficiency, and agricultural water assignment coefficients between 2024 and 2033. Residential area demand was found to be increasing with different rates under the scenarios between 47.40 and 73.01 sq.km. The SD-CA coupled framework presented in this research can be viewed as a decision support system to develop compensatory strategies for better management and planning of agricultural and residential lands.     

Efficient clustering in collaborative filtering recommender system: Hybrid method based on genetic algorithm and gravitational emulation local search algorithm

The ultimate goal of the Recommender System (RS) is to offer a proposal that is very close to the user's real opinion. Data clustering can be effective in increasing the accuracy of production proposals by the RS. In this paper, single-objective hybrid evolutionary approach is proposed for clustering items in the offline collaborative filtering RS. This method, after generating a population of randomized solutions, at each iteration, improves the population of solutions first by Genetic Algorithm (GA) and then by using the Gravitational Emulation Local Search (GELS) algorithm. Simulation results on standard datasets indicate that although the proposed hybrid meta-heuristic algorithm requires a relatively high run time, it can lead to more appropriate clustering of existing data and thus improvement of the Mean Absolute Error (MAE), Root Mean Square Error (RMSE) and Coverage criteria.     

城镇化过程中流域不透水面演变格局——以天津于桥水库流域为例

不透水面是衡量流域城镇化发展状况的一个关键指标,其扩展对流域生态水文过程产生重要影响.本文以天津于桥水库流域为例,在ENVI 5.1软件支持下,利用遥感影像获取1984、1994、2004和2013年4个时相的不透水面信息,采用修正后的归一化水体指数剔除水体信息,排除水体对不透水面提取精度的影响,运用线性光谱混合分析法生成流域不透水表面指数(ISA),并对其时空演变格局进行分析.结果表明: 模型均方根误差(RMSE)为0.005,像元精度为85.4%,试验结果准确可靠.1984—2013年,流域内不透水面覆盖度ISA平均值从0.16线性增长到0.23,在全流域范围内的不透水面面积增加了4.9%,其总不透水面面积增加近1倍.不透水面沿城镇区域呈现辐射式增长,沿子流域路网呈现填充式增长.不透水面覆盖度为中等等级的斑块形状不规则,破碎化程度最高.整个流域景观破碎化程度和多样性均逐年增高,人为干扰强度不断增大.     

基于Google Earth Engine与机器学习的大尺度30m分辨率沙地灌木覆盖度估算

相较于降雨充沛的南方,中国北方沙地植被呈现覆盖整体偏低、空间异质性强的特点。灌木作为该区域的优势植被,对于风沙固定、食品/木材供给起着极为重要的作用。针对当前大尺度、中高分辨率干旱地区灌木覆盖度遥感产品缺失的现状,研究提出了一套通过Collect Earth样本收集器进行样本采集、利用Google Earth Engine遥感云平台的数据与计算优势开展大尺度灌木覆盖度估算的方法,并选取中国北方四大沙地之一的毛乌素沙地开展了示范应用。研究结果表明:(1)Collect Earth样本收集器可以有效地获取地面灌木覆盖度样本数据集,可以将灌木与高大乔木与草本植被进行有效区分,为灌木覆盖度估算样本集的建立打下了基础;(2)利用Landsat数据与其他辅助数据,机器学习算法可以较好地实现灌木覆盖度的估算,CART模型确定性系数R~2为0.73,均方根误差(Root Mean Square Error, RMSE)为13.66%,预测精度(Estimated Accuracy, EA)为61.8%,SVM模型R~2为0.72,RMSE为13.73%,EA为61.6%;(3)提出的基于GEE的灌木覆盖度估算体系可为我国乃至全球尺度干旱地区沙地灌木覆盖度信息提取提供有效支撑,具有较大的应用潜力。     

Persistence and stability of macroinvertebrate communities in streams of Denali National Park, Alaska: implications for biological monitoring

1. Macroinvertebrate communities were studied from 1994 to 2001/2002 (except 1997) in six streams in Denali National Park, interior Alaska. All six streams were potential reference streams with no known impairment. 2. Abundance of individual taxa varied markedly from year to year. Overall, abundance decreased over the study period, particularly with respect to mayflies. Stonefly taxa showed lower persistence and were sometimes absent from a stream in any particular year. 3. Mean community persistence for the six streams, as measured by Jaccard's similarity coefficients between years, varied from 0.48 in the year pair 1999–2000 to 0.78 in 1998–99. Tattler Creek (a small stable stream) supported the most persistent macroinvertebrate community and Highway Pass Creek (a small, unstable creek) the least. Mean community persistence showed a significant relationship with mean winter snowfall (November to March) for the six streams. 4. The highest community compositional stability was found in Tattler Creek and the lowest in Highway Pass Creek, but stability varied markedly over time for the six streams, peaking in 1994–95 and reaching a minimum in 2000–01. Compositional stability was significantly related to the Pfankuch Index of channel stability. 5. The composition metrics % Chironomidae, % dominant taxa, %EPT, % Ephemeroptera and % Plecoptera, employed as part of the Alaska Stream Condition Index, varied over almost their entire range in these pristine streams across the 9 years of the study. 6. This study demonstrates the wide range of natural variation that occurs in benthic macroinvertebrate communities in these pristine central Alaskan streams, potentially limiting the applicability of composition metrics for the biological monitoring of water quality in these systems.     

流域景观格局与河流水质的多变量相关分析

流域内的景观格局改变是人类活动的宏观表现,会对河流水质产生显著影响,因此明确影响水质变化的关键景观因子,对于深入了解景观对水质的影响机制具有重要的研究价值。选择广东省淡水河流域为研究对象,以2007年ALOS卫星影像以及水质监测数据为基础,运用空间分析和多变量分析方法,分析淡水河流域景观格局与河流水质的相关关系。用包括流域和河岸带尺度的景观组成和空间结构信息的景观指数表征景观格局,用Spearman秩相关分析、多元线性逐步回归模型和典型相关分析(CCA)研究景观指数和水质指标的相关关系。研究结果表明:林地、城镇用地和农业用地占淡水河流域总面积超过90%,其中城镇用地超过20%。多元线性逐步回归分析和CCA结果说明水质指标受到多个景观指数的综合影响,反映了景观格局对水质的复杂影响机制。流域景观格局对河流水质有显著影响,流域尺度的景观指数比河岸带尺度的景观指数对水质影响更大。城镇用地比例是影响耗氧污染物和营养盐等污染物浓度最重要的景观指数,林地和农业用地对水质的影响较小。另外,景观破碎化对pH值、溶解氧和重金属等水质指标有显著影响。CCA的第一排序轴解释了景观指数与水质指标相关性的54.0%,前两排序轴累积能解释景观指数与水质指标相关性的87.6%,前两轴分别主要表达了城市化水平和景观破碎化水平的变化梯度。淡水河流域的景观格局特征从上游到下游呈现出城市—城乡交错—农村的景观梯度,水质变化也对应了这个梯度的变化,说明人类活动引起的流域土地覆盖及土地管理措施变化会对水质变化产生显著影响。     

Estimation and evaluation of pseudo-CT images using linear regression models and texture feature extraction from MRI images in the brain region to design external radiotherapy planning

AimThe aim of this study is to construct and evaluate Pseudo-CT images (P-CTs) for electron density calculation to facilitate external radiotherapy treatment planning.BackgroundDespite numerous benefits, computed tomography (CT) scan does not provide accurate information on soft tissue contrast, which often makes it difficult to precisely differentiate target tissues from the organs at risk and determine the tumor volume. Therefore, MRI imaging can reduce the variability of results when registering with a CT scan.Materials and methodsIn this research, a fuzzy clustering algorithm was used to segment images into different tissues, also linear regression methods were used to design the regression model based on the feature extraction method and the brightness intensity values. The results of the proposed algorithm for dose-volume histogram (DVH), Isodose curves, and gamma analysis were investigated using the RayPlan treatment planning system, and VeriSoft software. Furthermore, various statistical indices such as Mean Absolute Error (MAE), Mean Error (ME), and Structural Similarity Index (SSIM) were calculated.ResultsThe MAE of a range of 45–55 was found from the proposed methods. The relative difference error between the PTV region of the CT and the Pseudo-CT was 0.5, and the best gamma rate was 95.4% based on the polar coordinate feature and proposed polynomial regression model.ConclusionThe proposed method could support the generation of P-CT data for different parts of the brain region from a collection of MRI series with an acceptable average error rate by different evaluation criteria.     

Assessing historical fish community composition using surveys, historical collection data, and species distribution models

Accurate establishment of baseline conditions is critical to successful management and habitat restoration. We demonstrate the ability to robustly estimate historical fish community composition and assess the current status of the urbanized Barton Creek watershed in central Texas, U.S.A. Fish species were surveyed in 2008 and the resulting data compared to three sources of fish occurrence information: (i) historical records from a museum specimen database and literature searches; (ii) a nearly identical survey conducted 15 years earlier; and (iii) a modeled historical community constructed with species distribution models (SDMs). This holistic approach, and especially the application of SDMs, allowed us to discover that the fish community in Barton Creek was more diverse than the historical data and survey methods alone indicated. Sixteen native species with high modeled probability of occurrence within the watershed were not found in the 2008 survey, seven of these were not found in either survey or in any of the historical collection records. Our approach allowed us to more rigorously establish the true baseline for the pre-development fish fauna and then to more accurately assess trends and develop hypotheses regarding factors driving current fish community composition to better inform management decisions and future restoration efforts. Smaller, urbanized freshwater systems, like Barton Creek, typically have a relatively poor historical biodiversity inventory coupled with long histories of alteration, and thus there is a propensity for land managers and researchers to apply inaccurate baseline standards. Our methods provide a way around that limitation by using SDMs derived from larger and richer biodiversity databases of a broader geographic scope. Broadly applied, we propose that this technique has potential to overcome limitations of popular bioassessment metrics (e.g., IBI) to become a versatile and robust management tool for determining status of freshwater biotic communities.     

Estimating Carbon Flux Phenology with Satellite-Derived Land Surface Phenology and Climate Drivers for Different Biomes: A Synthesis of AmeriFlux Observations

Carbon Flux Phenology (CFP) can affect the interannual variation in Net Ecosystem Exchange (NEE) of carbon between terrestrial ecosystems and the atmosphere. In this study, we proposed a methodology to estimate CFP metrics with satellite-derived Land Surface Phenology (LSP) metrics and climate drivers for 4 biomes (i.e., deciduous broadleaf forest, evergreen needleleaf forest, grasslands and croplands), using 159 site-years of NEE and climate data from 32 AmeriFlux sites and MODIS vegetation index time-series data. LSP metrics combined with optimal climate drivers can explain the variability in Start of Carbon Uptake (SCU) by more than 70% and End of Carbon Uptake (ECU) by more than 60%. The Root Mean Square Error (RMSE) of the estimations was within 8.5 days for both SCU and ECU. The estimation performance for this methodology was primarily dependent on the optimal combination of the LSP retrieval methods, the explanatory climate drivers, the biome types, and the specific CFP metric. This methodology has a potential for allowing extrapolation of CFP metrics for biomes with a distinct and detectable seasonal cycle over large areas, based on synoptic multi-temporal optical satellite data and climate data.     

Comparability of a regional and state survey: effects on fish IBI assessment for West Virginia,U.S.A.

Probability-based survey designs are now being investigated to allow condition to be assessed for a discrete population of watershed management units and to infer probability of impairment to other unsampled watersheds. Results can be used to focus further monitoring and restoration efforts. Fish community data and index of biotic integrity (IBI) development were compared between the 1993 and 1998 Environmental Monitoring and Assessment Program Mid-Atlantic Integrated Assessment (EMAP-MAIA) survey and a West Virginia Regional EMAP (WV REMAP) survey conducted in 2001–2002. Both designs were based on probability surveys, but the EMAP design treated streams as a continuous linear network comprising an infinite population of points, while the REMAP design used a discrete set of watershed outlets as defined by 12-digit Hydrologic Cataloging Units (HUC12) as the sample population. The comparability of the watershed-based WV REMAP survey design results with the linear network-based EMAP-MAIA survey results for West Virginia was affected by the different size range of watershed areas included in each target population. Once similar watershed area ranges were considered by narrowing the size range included in the West Virginia EMAP-MAIA data set, virtually identical cumulative distribution functions for fish IBI scores were obtained. The reduced variability in reference conditions obtained by applying a restricted range of watershed areas allowed us to detect and correct for ecoregional differences in fish IBI metrics and scores, after excluding the biogeographically distinct Potomac River drainage basin located in the Central Appalachian Ridge and Valley Ecoregion. Handling editor: K. Martens     

A No-Reference Quality Metric for Parameter Tuning of Edge-Aware Filters – An Anti-Image Forensic Method

Edge-Aware Filters (EAF) are less frequently detected by forensic tools compared to the median filter. However, EAFs also blur the edges, if their operational parameters are not tuned properly. Objective image quality metrics which reflect the quality of the smoothed images are necessary for tuning the operational parameters. A novel formulation of a no-reference composite metric, termed as Denoising Performance Metric (DPM) is introduced in this paper. DPM exhibited a correlation of 0.98 ± 0.02 and 0.89 ± 0.09 with the Mean Opinion Score (MOS) and Peak Signal to Noise Ratio (PSNR) between the test images and the noise-free benchmark image, respectively. The correlation observed for type-2 Vector Mean Squared Error (VMSE) with MOS and PSNR are 0.97 ± 0.02 and 0.79 ± 0.25, respectively. The proposed metric is observed to be superior to type-2 Vector Mean Squared Error (VMSE) in terms of its correlation with the subjective fidelity ratings. It can be used for tuning operational parameters of EAF to enhance their ability to tackle forensic tools.     

城市地表温度与关键景观要素的关系

利用Landsat ETM+遥感影像,提取上海市外环线范围内的地表温度、不透水面率、归一化差值植被指数、改进的归一化差异水体指数,定量研究地表温度与城市关键景观类型之间的关系.结果表明:地表温度与不透水面率呈显著的线性正相关( R2=0.837);地表温度与归一化差值植被指数和改进的归一化差异水体指数呈非线性关系,但地表温度与正的归一化差值植被指数和正的改进的归一化差值水体指数呈显著线性关系.鉴于归一化差值植被指数和改进的归一化差异水体指数大于0时才能真正代表植被和水体,因此,建议今后研究地表温度时使用正的归一化差值植被指数和改进的归一化差异水体指数;地表温度与不透水面率、归一化差值植被指数和改进的归一化差值水体指数的多元线性回归分析表明,不透水面起着增温作用,植被、水体起降温作用,植被较水体的降温作用大.     

Comparison of watershed disturbance predictive models for stream benthic macroinvertebrates for three distinct ecoregions in western US

The successful use of macroinvertebrates as indicators of stream condition in bioassessments has led to heightened interest throughout the scientific community in the prediction of stream condition. For example, predictive models are increasingly being developed that use measures of watershed disturbance, including urban and agricultural land-use, as explanatory variables to predict various metrics of biological condition such as richness, tolerance, percent predators, index of biotic integrity, functional species traits, or even ordination axes scores. Our primary intent was to determine if effective models could be developed using watershed characteristics of disturbance to predict macroinvertebrate metrics among disparate and widely separated ecoregions. We aggregated macroinvertebrate data from universities and state and federal agencies in order to assemble stream data sets of high enough density appropriate for modeling in three distinct ecoregions in Oregon and California. Extensive review and quality assurance of macroinvertebrate sampling protocols, laboratory subsample counts and taxonomic resolution was completed to assure data comparability. We used widely available digital coverages of land-use and land-cover data summarized at the watershed and riparian scale as explanatory variables to predict macroinvertebrate metrics commonly used by state resource managers to assess stream condition. The “best” multiple linear regression models from each region required only two or three explanatory variables to model macroinvertebrate metrics and explained 41–74% of the variation. In each region the best model contained some measure of urban and/or agricultural land-use, yet often the model was improved by including a natural explanatory variable such as mean annual precipitation or mean watershed slope. Two macroinvertebrate metrics were common among all three regions, the metric that summarizes the richness of tolerant macroinvertebrates (RICHTOL) and some form of EPT (Ephemeroptera, Plecoptera, and Trichoptera) richness. Best models were developed for the same two invertebrate metrics even though the geographic regions reflect distinct differences in precipitation, geology, elevation, slope, population density, and land-use. With further development, models like these can be used to elicit better causal linkages to stream biological attributes or condition and can be used by researchers or managers to predict biological indicators of stream condition at unsampled sites.     

Deep learning dose prediction for IMRT of esophageal cancer: The effect of data quality and quantity on model performance

PurposeTo investigate the effect of data quality and quantity on the performance of deep learning (DL) models, for dose prediction of intensity-modulated radiotherapy (IMRT) of esophageal cancer.Material and methodsTwo databases were used: a variable database (VarDB) with 56 clinical cases extracted retrospectively, including user-dependent variability in delineation and planning, different machines and beam configurations; and a homogenized database (HomDB), created to reduce this variability by re-contouring and re-planning all patients with a fixed class-solution protocol.Experiment 1 analysed the user-dependent variability, using 26 patients planned with the same machine and beam setup (E26-VarDB versus E26-HomDB). Experiment 2 increased the training set by groups of 10 patients (E16, E26, E36, E46, and E56) for both databases.Model evaluation metrics were the mean absolute error (MAE) for selected dose-volume metrics and the global MAE for all body voxels.ResultsFor Experiment 1, E26-HomDB reduced the MAE for the considered dose-volume metrics compared to E26-VarDB (e.g. reduction of 0.2 Gy for D95-PTV, 1.2 Gy for Dmean-heart or 3.3% for V5-lungs). For Experiment 2, increasing the database size slightly improved performance for HomDB models (e.g. decrease in global MAE of 0.13 Gy for E56-HomDB versus E26-HomDB), but increased the error for the VarDB models (e.g. increase in global MAE of 0.20 Gy for E56-VarDB versus E26-VarDB).ConclusionA small database may suffice to obtain good DL prediction performance, provided that homogenous training data is used. Data variability reduces the performance of DL models, which is further pronounced when increasing the training set.     

Decadal‐scale aspen changes: evidence in remote sensing and tree ring data

Question: (1) Which remote sensing classification most successfully identify aspen using multitemporal Landsat 5 TM images and airborne lidar data? (2) How has aspen distribution changed in southwestern Idaho? (3) Are topographic variables and conifer encroachment correlated with aspen changes? Location: Reynolds Creek Experimental Watershed in southwestern Idaho, USA. Methods: Multi‐temporal Landsat 5 TM and lidar data were used individually and fused together. The best classification model was compared with a 1965 aspen map and tree ring data. Conifer encroachment was examined via image‐based change detection and field mapping. Lidar‐derived topographic variables were correlated with aspen change patterns using quantile regression models. Results: The best Landsat 5 TM classification was a normalized difference vegetation index (NDVI)‐based approach with 92% overall accuracy. The lidar classification of tree presence/absence performed with 100% overall accuracy. Fusing the lidar classification with various Landsat 5 TM classifications improved overall accuracies 3 to 6%. Among the fusion models, the NDVI‐lidar fusion performed best with 96% overall accuracy. Change detection indicated 69% decline in aspen cover, but 179% increase in aspen cover in other areas of the watershed. Conifers have completely replaced 17% of the aspen, while 93% of the remaining aspen stands have young Douglas‐fir and western juniper trees underneath the aspen canopy. Aspen significantly decreased (P‐values <0.05) with increasing elevation (up to 2150 m) and decreasing slope. Conclusions: Landsat 5 TM data used with a NDVI‐based approach provide an accurate method to classify aspen distribution. Landsat 5 TM classifications can be further improved via fusion with lidar data. Aspen change patterns are spatially variable: while aspen is drastically declining in some parts of this watershed, aspen is increasing in other areas.