Towards the development of a regional version of MOD17 for the determination of gross and net primary productivity of oil palm trees

Conducting quantitative studies on the carbon balance or productivity of oil palm is important for understanding the role of this ecosystem in global climate change. The MOD17 algorithm is used for processing data from the Moderate Resolution Imaging Spectroradiometer (MODIS) to generate the values of gross primary productivity (GPP) and net primary productivity for input to global carbon cycle modelling. In view of the increasing importance of data on carbon sequestration at regional and national levels, we have studied one important factor affecting the accuracy of the implementation of MOD17 at the sub-global level, namely the database of MODIS land cover (MOD12Q1) used by MOD17. By using a study area of approximately 7 km × 7 km (49 MODIS pixels) in semi-rural Johor in Peninsular Malaysia and using Google Earth 0.75 m resolution images as ground data, we found that the land-cover type for only 16 of these 49 MODIS pixels was correctly identified by MOD12Q1 using its 1 km resolution land-cover database. This leads to errors of 24% to 50% in the maximum light use efficiency, leading to corresponding errors of 24% to 50% in the GPP. We show that by using the Finer Resolution Observation and Monitoring – Global Land Cover (FROM-GLC) land-cover database developed by Gong et al., this particular error can be essentially eliminated, but at the cost of using extra computing resources.     

On the upstream inputs into the MODIS primary productivity products using biometric data from oil palm plantations

This study evaluated the influence of upstream inputs into the Moderate Resolution Imaging Spectroradiometer (MODIS) primary productivity products, termed the MOD17, at tropical oil palm plantations (Elaeis guineensis Jacq.). Evaluation of MOD17 using oil palm plantations as test sites is ideal because the plantations are cultivated on large areas which are comparable with the size of MODIS pixels. It is difficult to find test sites covered by other single species in a whole pixel. The upstream inputs studied included (1) MODIS land cover, (2) the National Centers for Environmental Prediction–Department of Energy (NCEP-DOE) Reanalysis 2 meteorological data set, (3) MODIS leaf area index/fraction of photosynthetically active radiation (LAI/fPAR), and (4) MODIS maximum light-use efficiency (maximum LUE). Oil palm biometric and local meteorological data were utilized as ground data. Furthermore, scaling up oil palm LAI and fPAR from plot scale to regional scale (Peninsular Malaysia) was done empirically by correlating oil palm LAI derived from the hemispherical photography technique with radiance information from the Disaster Monitoring Constellation 2 satellite (UK-DMC 2). The upscaled LAI/fPAR developed in this study was used to evaluate the MODIS LAI/fPAR. The results showed that the MODIS land-cover product has an overall accuracy of 78.8% when compared to the Peninsular Malaysia land-use map produced by the Department of Agriculture, Malaysia. Regarding the NCEP-DOE Reanalysis 2 data set, vapour pressure deficit (VPD) and photosynthetically active radiation (PAR) contain large uncertainties in our study area. However, MODIS LAI and fPAR were correlated relatively well with the upscaled LAI (R² = 0.50) and the upscaled fPAR (R² = 0.60), respectively. The constant values of maximum LUE for croplands and evergreen broadleaf forest ecosystems are lower than the maximum LUE of oil palm. The relative predictive error assessment showed that the MOD17 net primary productivity (NPP) overestimated oil palm NPP derived from biometric methods by 142–204%. We replaced the upstream inputs of MOD17 by the local inputs for estimating oil palm GPP and NPP in Peninsular Malaysia. This was done by (1) assigning maximum LUE for oil palm plantations as a constant at 1.68 g C m^?2 day^?1, (2) utilizing meteorological data from local meteorological stations, and (3) using the upscaled fPAR of oil palm plantations. The amount of oil palm GPP and NPP for Peninsular Malaysia in 2010 were estimated to be ~0.09 Pg C year^?1 (or equivalent to ~0.33 Pg CO₂ year^?1) and ~0.03 Pg C year^?1 (~0.11 Pg CO₂ year^?1), respectively, indicating that oil palm plantations in Peninsular Malaysia can play an important role in global carbon sequestration. In the future there is likely to be a demand for MODIS GPP and NPP products that are more accurate than those currently generated by MOD17. We recommend future developments of the MOD17 processing system to allow improvements in the upstream input parameters, in the manner described in this article, both for global processing and for the production of more accurate values for GPP and NPP at regional and local scales.     

AMSR-E亮温数据与MODIS陆表分类产品结合反演全球陆表温度

AMSR-E被动微波传感器获取的亮温数据与MODIS陆表分类产品(MOD12)相结合,将全球陆表分为16类,并假设每种类型的地表在各个被动微波通道具有较一致的发射率,在此基础上针对每种陆表类型分别建立了陆表温度反演算法。在算法的建立过程中,为了避免混合像元以及冻土、积雪发射率不确定性带来的影响,仅对单一地表类型占90%以上以及MODIS陆表温度产品高于273K的被动微波像元进行回归。同时,考虑到降雨对回归结果的影响,在数据选择中加入了降雨判识,在被动微波亮温数据中除去了降雨像元。利用上述算法,用2004年1～10月的全球部分地区AMSR-E数据在MODIS陆表分类产品的基础上对每种地表类型分别进行了陆表温度反演,并与MODIS陆表温度产品进行对比,结果显示相关性较好,均方根误差为2～4 K。     

Use of pixel- and plot-scale screening variables to validate MODIS GPP predictions with Forest Inventory and Analysis NPP measures across the eastern USA

Moderate Resolution Imaging Spectroradiometer (MODIS) estimates of gross primary production (GPP) were validated using field-based estimates of net primary production from the Forest Inventory and Analysis (FIA) Program across the eastern USA. A total of 54 969 MODIS pixels and co-located FIA plots were analysed to validate MODIS GPP estimates. We used a data resolution of individual MODIS pixels and co-located FIA plots, and used detailed pixel- and plot-specific attributes by applying screening variables (SVs) to assess conditions under which MODIS GPP was most strongly validated. Eight SVs were used to test six hypotheses about the conditions under which MODIS GPP would be most strongly validated. The six hypotheses addressed were (1) MODIS pixel quality checks, (2) FIA plot quality checks, (3) land-cover classification comparability of co-located MODIS pixels and FIA plots, (4) FIA plot homogeneity, (5) FIA plot tree density and (6) MODIS seasonal variation. SVs were assessed in terms of trade-off between improved relations and reduced number of samples. MODIS seasonal variation and FIA plot tree density were the two most efficient SVs, followed by basic quality checks for each data set. Sequential application of SVs indicated that combined usage of five of the eight SVs provided an efficient data set of 17 090 co-located MODIS pixels and FIA plots, which raised the Pearson correlation coefficient from 0.01 for the Complete data set of 54 969 plots to 0.48 for this screened subset of 17 090 plots. The screened subset of plots exhibited good representation of the Complete data set in terms of species abundance, plot distribution and mean productivity. We conclude that the application of SVs provides a useful approach to ensure compatibility of two data sets for broad-scale forest carbon budget analysis and monitoring.     

Downscaling MODIS LST in the East African mountains using elevation gradient and land-cover information

Land-surface temperature (LST) is strongly affected by altitude and surface albedo. In mountain regions where steep slopes and heterogeneous land cover are predominant, LST can vary significantly within short distances. Although remote sensing currently provides opportunities for monitoring LST in inaccessible regions, the coarse resolution of some sensors may result in large uncertainties at sub-pixel scales. This study aimed to develop a simple methodology for downscaling 1 km Moderate Resolution Spectroradiometer (MODIS) LST pixels, by accounting for sub-pixel LST variation associated with altitude and land-cover spatial changes. The approach was tested in Mount Kilimanjaro, Tanzania, where changes in altitude and vegetation can take place over short distances. Daytime and night-time MODIS LST estimates were considered separately. A digital elevation model (DEM) and normalized difference vegetation index (NDVI), both at 250 m spatial resolution, were used to assess altitude and land-cover changes, respectively. Simple linear regressions and multivariate regressions were used to quantify the relationship between LST and the independent variables, altitude and NDVI. The results show that, in Kilimanjaro, altitude variation within the area covered by a 1 km MODIS LST pixel can be up to ±300 m. These altitude changes can cause sub-pixel variation of up to ±2.13°C for night-time and ±2.88°C for daytime LST. NDVI variation within 1 km pixels ranged between –0.2 and 0.2. For night-time measurements, altitude explained up to 97% of LST variation, while daytime LST was strongly affected by land cover. Using multivariate regressions, the combination of altitude and NDVI explained up to 94% of daytime LST variation in Kilimanjaro. Finally, the downscaling approach proposed in this study allowed an improved representation of the influence of landscape features on local-scale LST patterns.     

基于涡度数据的东北草地光能利用率模型构建与验证

草地作为地球上分布最广的植被类型,在陆地碳循环中发挥着重要作用。草地生产力是估算产草量的基础,准确模拟生产力对草原资源合理利用及生态保护具有重要意义。以东北草地生产力为研究核心,利用涡度相关通量观测数据、遥感数据和气象数据,构建和检验东北草地光能利用率模型。东北草地光能利用率模型以归一化物候植被指数(NDPI)代表光合有效辐射吸收比例,以地表水分指数(LSWI)+0.5表示水分胁迫因子。基于44个草原站的通量数据对东北草地光能利用率模型进行验证,东北草地光能利用率模型的R²为0.855,高于MODIS GPP产品(R²=0.719),略高于VPM GPP产品(R²=0.848),东北草地光能利用率模型的MAE和RMSE分别为0.374 g Cm^-2和0.735 g Cm^-2,低于MODIS GPP产品(MAE=0.562 g Cm^-2,RMSE=1.026 g Cm^-2)和VPM GPP产品(MAE=0.667 g Cm^-2...     

Effects of forest type and environmental factors on forest carbon use efficiency assessed using MODIS and FIA data across the eastern USA

The carbon use efficiency (CUE) of a forest, calculated as the ratio of net primary productivity (NPP) to gross primary productivity (GPP), measures how efficiently a forest sequesters atmospheric carbon. Some prior research has suggested that CUE varies with environmental conditions, while other suggests that CUE is constant. Research using Moderate Resolution Imaging Spectroradiometer (MODIS) data has indicated a variable CUE, but those results are suspected because MODIS NPP data have not been well validated.

We tested two questions. First, whether MODIS CUE is constant or whether it varies by forest type, climate, and geographic factors across the eastern USA. Second, whether those results occur when field-based NPP data are employed. We used MODIS model-based estimates of GPP and NPP, and forest inventory and anlaysis (FIA) field-based estimates of NPP data. We calculated two estimates of CUE for forest in 390 km² hexagons: (1) MODIS CUE as MODIS NPP divided by MODIS GPP and (2) F/M Z_CUE as the standardized difference between FIA NPP and MODIS GPP.

MODIS CUE and F/M Z_CUE both varied similarly and significantly in relation to forest type, and climatic and geographic factors, strongly supporting a variable rather than a constant CUE. The CUE was significantly higher in deciduous than in mixed and evergreen forests. Regression models indicated that CUE decreased with increases in temperature and precipitation and increased with latitude and altitude. The similar trends in MODIS CUE and F/M Z_CUE support the use of the more easily obtained MODIS CUE.     

Scaling and assessment of GPP from MODIS using a combination of airborne lidar and eddy covariance measurements over jack pine forests

Understanding the influence of within-pixel land cover heterogeneity is essential for the extrapolation of measured and modeled CO₂ fluxes from the canopy to regional scales using remote sensing. Airborne light detection and ranging (lidar) was used to estimate spatial and temporal variations of gross primary production (GPP) across a jack pine chronosequence of four sites in Saskatchewan, Canada for comparison with the Moderate Resolution Imaging Spectroradiometer (MODIS) GPP product. This study utilizes high resolution canopy structural information obtained from airborne lidar to bridge gaps in spatial representation between plot, eddy covariance (EC), and MODIS estimates of vegetation GPP. First we investigate linkages between canopy structure obtained from measurements and light response curves at a jack pine chronosequence during the growing season of 2004. Second, we use the measured canopy height and foliage cover inputs to create a structure-based GPP model (GPP_Landsberg) which was tested in 2005. The GPP model is then run using lidar data (GPP_Lidar) and compared with eight-day cumulative MODIS GPP (GPP_MODIS) and EC observations (GPP_EC). Finally, we apply the lidar GPP model at spatial resolutions of 1 m to 1000 m to examine the influence of within-pixel heterogeneity and scaling (or pixel aggregation) on GPP_Lidar. When compared over eight-day cumulative periods throughout the 2005 growing season, the standard deviation of differences between GPP_lidar and GPP_MODIS were less than differences between either of them and GPP_EC at all sites. As might be expected, the differences between pixel aggregated GPP estimates are most pronounced at sites with the highest levels of spatial canopy heterogeneity. The results of this study demonstrate one method for using lidar to scale between eddy covariance flux towers and coarse resolution remote sensing pixels using a structure-based Landsberg light curve model.     

Mapping regional land cover with MODIS data for biological conservation: Examples from the Greater Yellowstone Ecosystem, USA and Pará State, Brazil

The paper investigated the application of MODIS data for mapping regional land cover at moderate resolutions (250 and 500 m), for regional conservation purposes. Land cover maps were generated for two major conservation areas (Greater Yellowstone Ecosystem—GYE, USA and the Pará State, Brazil) using MODIS data and decision tree classifications. The MODIS land cover products were evaluated using existing Landsat TM land cover maps as reference data. The Landsat TM land cover maps were processed to their fractional composition at the MODIS resolution (250 and 500 m). In GYE, the MODIS land cover was very successful at mapping extensive cover types (e.g. coniferous forest and grasslands) and far less successful at mapping smaller habitats (e.g. wetlands, deciduous tree cover) that typically occur in patches that are smaller than the MODIS pixels, but are reported to be very important to biodiversity conservation. The MODIS classification for Pará State was successful at producing a regional forest/non-forest product which is useful for monitoring the extreme human impacts such as deforestation. The ability of MODIS data to map secondary forest remains to be tested, since regrowth typically harbors reduced levels of biodiversity. The two case studies showed the value of using multi-date 250 m data with only two spectral bands, as well as single day 500 m data with seven spectral bands, thus illustrating the versatile use of MODIS data in two contrasting environments. MODIS data provide new options for regional land cover mapping that are less labor-intensive than Landsat and have higher resolution than previous 1 km AVHRR or the current 1 km global land cover product. The usefulness of the MODIS data in addressing biodiversity conservation questions will ultimately depend upon the patch sizes of important habitats and the land cover transformations that threaten them.     

Global land cover classification at 1 km spatial resolution using a classification tree approach

This paper on reports the production of a 1 km spatial resolution land cover classification using data for 1992-1993 from the Advanced Very High Resolution Radiometer (AVHRR). This map will be included as an at-launch product of the Moderate Resolution Imaging Spectroradiometer (MODIS) to serve as an input for several algorithms requiring knowledge of land cover type. The methodology was derived from a similar effort to create a product at 8 km spatial resolution, where high resolution data sets were interpreted in order to derive a coarse-resolution training data set. A set of 37 294 x 1 km pixels was used within a hierarchical tree structure to classify the AVHRR data into 12 classes. The approach taken involved a hierarchy of pair-wise class trees where a logic based on vegetation form was applied until all classes were depicted. Multitemporal AVHRR metrics were used to predict class memberships. Minimum annual red reflectance, peak annual Normalized Difference Vegetation Index (NDVI), and minimum channel three brightness temperature were among the most used metrics. Depictions of forests and woodlands, and areas of mechanized agriculture are in general agreement with other sources of information, while classes such as low biomass agriculture and high-latitude broadleaf forest are not. Comparisons of the final product with regional digital land cover maps derived from high-resolution remotely sensed data reveal general agreement, except for apparently poor depictions of temperate pastures within areas of agriculture. Distinguishing between forest and non-forest was achieved with agreements ranging from 81 to 92% for these regional subsets. The agreements for all classes varied from an average of 65% when viewing all pixels to an average of 82% when viewing only those 1 km pixels consisting of greater than 90% one class within the high-resolution data sets.     

Evaluation of Collections 4 and 5 of the MODIS Gross Primary Productivity product and algorithm improvement at a tropical savanna site in northern Australia

In this study, we assessed the accuracy of the MODIS (Moderate Resolution Imaging Spectroradiometer) GPP (gross primary productivity) Collections 4.5, 4.8 and 5 along with Leaf Area Index (LAI), fraction of absorbed Photosynthetically Active Radiation (fPAR), light use efficiency (LUE) and meteorological variables that are used to estimate GPP for a northern Australian savanna site. Results of this study indicated that the MODIS products captured the seasonal variation in GPP, LAI and fPAR well. Using the index of agreement (IOA), it was found that Collections 4.5 and 4.8 (IOA 0.89 respectively) agreed reasonably well with flux tower measurements between 2001 and 2006. It was also found that MODIS Collection 4.5 predicted the dry season GPP well (Relative Predictive Error (RPE) 4.17%, IOA 0.72 and Root Mean Square Error (RMSE) of 1.05 g C m^− 2 day^− 1), whilst Collection 4.8 performed better in capturing wet season dynamics (RPE 1.11%, IOA 0.80 and RMSE of 0.91 g C m^− 2 day^− 1). Although the wet season magnitude of GPP was predicted well by Collection 4.8, an examination of the inputs to the GPP algorithm revealed that MODIS fPAR was too high, but this was compensated by PAR and LUE that was too low. Although LAI and fPAR estimated by Collection 5 were more accurate, GPP for this Collection resulted in a much lower value (RPE 25%) due to errors in other factors. Recalculation of MODIS GPP using site specific input parameters indicated that MODIS fPAR was the main reason for the differences between MODIS and tower derived GPP followed by LUE and meteorological inputs. GPP calculated using all site specific values agreed very well with tower data on an annual basis (IOA 0.94, RPE 6.06% and RMSE 0.83 g C m^− 2 day^− 1) but the early initiation of the growing season calculated by the MODIS algorithm was improved when the vapor pressure deficit (VPD) function was replaced with a soil water deficit function. The results of this study however, reinforce previous findings in water limited regions, like Australia, and incorporation of soil moisture in a LUE model is needed to accurately estimate the productivity.     

Fractional snow cover mapping through artificial neural network analysis of MODIS surface reflectance

Accurate areal measurements of snow cover extent are important for hydrological and climate modeling. The traditional method of mapping snow cover is binary where a pixel is considered either snow-covered or snow-free. Fractional snow cover (FSC) mapping can achieve a more precise estimate of areal snow cover extent by estimating the fraction of a pixel that is snow-covered. The most common snow fraction methods applied to Moderate Resolution Imaging Spectroradiometer (MODIS) images have been spectral unmixing and an empirical Normalized Difference Snow Index (NDSI). Machine learning is an alternative for estimating FSC as artificial neural networks (ANNs) have been successfully used for estimating the subpixel abundances of other surfaces. The advantages of ANNs are that they can easily incorporate auxiliary information such as land cover type and are capable of learning nonlinear relationships between surface reflectance and snow fraction. ANNs are especially applicable to mapping snow cover extent in forested areas where spatial mixing of surface components is nonlinear. This study developed a multilayer feed-forward ANN trained through backpropagation to estimate FSC using MODIS surface reflectance, NDSI, Normalized Difference Vegetation Index (NDVI) and land cover as inputs. The ANN was trained and validated with higher spatial-resolution FSC maps derived from Landsat Enhanced Thematic Mapper Plus (ETM+) binary snow cover maps. Testing of the network was accomplished over training and independent test areas. The developed network performed adequately with RMSE of 12% over training areas and slightly less accurately over the independent test scenes with RMSE of 14%. The developed ANN also compared favorably to the standard MODIS FSC product. The study also presents a comprehensive validation of the standard MODIS snow fraction product whose performance was found to be similar to that of the ANN.     

Sub-pixel estimation of urban land cover components with linear mixture model analysis and Landsat Thematic Mapper imagery

We examine the utility of linear mixture modelling in the sub-pixel analysis of Landsat Enhanced Thematic Mapper (ETM) imagery to estimate the three key land cover components in an urban/suburban setting: impervious surface, managed/unmanaged lawn and tree cover. The relative effectiveness of two different endmember sets was also compared. The interior endmember set consisted of the median pixel value of the training pixels of each land cover and the exterior endmember set was the extreme pixel value. As a means of accuracy assessment, the resulting land cover estimates were compared with independent estimates obtained from the visual interpretation of digital orthophotography and classified IKONOS imagery. Impervious surface estimates from the Landsat ETM showed a high degree of similarity (RMS error (RMSE) within approximately ±10 to 15%) to that obtained using high spatial resolution digital orthophotography and IKONOS imagery. The partition of the vegetation component into tree vs grass cover was more problematic due to the greater spectral similarity between these land cover types with RMSE of approximately ±12 to 22%. The interior endmember set appeared to provide better differentiation between grass and urban tree cover than the exterior endmember set. The ability to separate the grass vs tree components in urban vegetation is of major importance to the study of the urban/suburban ecosystems as well as watershed assessment.     

Improvements of the MODIS terrestrial gross and net primary production global data set

MODIS primary production products (MOD17) are the first regular, near-real-time data sets for repeated monitoring of vegetation primary production on vegetated land at 1-km resolution at an 8-day interval. But both the inconsistent spatial resolution between the gridded meteorological data and MODIS pixels, and the cloud-contaminated MODIS FPAR/LAI (MOD15A2) retrievals can introduce considerable errors to Collection4 primary production (denoted as C4 MOD17) results. Here, we aim to rectify these problems through reprocessing key inputs to MODIS primary vegetation productivity algorithm, resulting in improved Collection5 MOD17 (here denoted as C5 MOD17) estimates. This was accomplished by spatial interpolation of the coarse resolution meteorological data input and with temporal filling of cloud-contaminated MOD15A2 data. Furthermore, we modified the Biome Parameter Look-Up Table (BPLUT) based on recent synthesized NPP data and some observed GPP derived from some flux tower measurements to keep up with the improvements in upstream inputs. Because MOD17 is one of the down-stream MODIS land products, the performance of the algorithm can be largely influenced by the uncertainties from upstream inputs, such as land cover, FPAR/LAI, the meteorological data, and algorithm itself. MODIS GPP fits well with GPP derived from 12 flux towers over North America. Globally, the 3-year MOD17 NPP is comparable to the Ecosystem Model-Data Intercomparison (EMDI) NPP data set, and global total MODIS GPP and NPP are inversely related to the observed atmospheric CO₂ growth rates, and MEI index, indicating MOD17 are reliable products. From 2001 to 2003, mean global total GPP and NPP estimated by MODIS are 109.29 Pg C/year and 56.02 Pg C/year, respectively. Based on this research, the improved global MODIS primary production data set is now ready for monitoring ecological conditions, natural resources and environmental changes.     

Improving continuity of MODIS terrestrial photosynthesis products using an interpolation scheme for cloudy pixels

The Moderate Imaging Spectroradiometer (MODIS) sensors onboard the NASA Terra and Aqua satellites provide the means for frequent measurement and monitoring of the status and seasonal variability in global vegetation phenology and productivity. However, while MODIS reflectance data are often interrupted by clouds, terrestrial processes like photosynthesis are continuous, so MODIS photosynthesis data must be able to cope with cloudy pixels. We developed cloud‐correction algorithms to improve retrievals of the MODIS photosynthesis product (PSNnet) corresponding to clear sky conditions by proposing four alternative cloud‐correction algorithms, which have different levels of complexity and correct errors associated with cloudy‐pixel surface reflectance. The cloud‐correction algorithms were applied at four weather stations, two fluxtower sites and the Pacific Northwest (PNW) region of the USA to test a range of cloud climatologies. Application of the cloud‐correction algorithms increased the magnitude of both daily and annual MODIS PSNnet results. Our results indicate that the proposed cloud correction methods improve the current MODIS PSNnet product considerably at both site and regional scales and weekly to annual time steps for areas subjected to frequent cloud cover. The corrections can be applied as a post‐processing interpolation of PSNnet, and do not require reprocessing of the MOD17A2 algorithm.     

Estimating northern peatland CO2 exchange from MODIS time series data

Studies using satellite sensor-derived data as input to models for CO₂ exchange show promising results for closed forest stands. There is a need for extending this approach to other land cover types, in order to carry out large-scale monitoring of CO₂ exchange. In this study, three years of eddy covariance data from two peatlands in Sweden were averaged for 16-day composite periods and related to data from the Moderate Resolution Imaging Spectroradiometer (MODIS) and modeled photosynthetic photon flux density (PPFD). Noise in the time series of MODIS 250 m vegetation indices was reduced by using double logistic curve fits. Smoothed normalized difference vegetation index (NDVI) showed saturation during summertime, and the enhanced vegetation index (EVI) generally gave better results in explaining gross primary productivity (GPP). The strong linear relationships found between GPP and the product of EVI and modeled PPFD (R² = 0.85 and 0.76) were only slightly stronger than for the product of EVI and MODIS daytime 1 km land surface temperature (LST) (R² = 0.84 and 0.71). One probable reason for these results is that several controls on GPP were related to both modeled PPFD and daytime LST. Since ecosystem respiration (ER) was largely explained by diurnal LST in exponential relationships (R² = 0.89 and 0.83), net ecosystem exchange (NEE) was directly related to diurnal LST in combination with the product of EVI and modeled PPFD in multiple exponential regressions (R² = 0.81 and 0.73). Even though the R² values were somewhat weaker for NEE, compared to GPP and ER, the RMSE values were much lower than if NEE would have been estimated as the sum of GPP and ER. The overall conclusion of this study is that regression models driven by satellite sensor-derived data and modeled PPFD can be used to estimate CO₂ fluxes in peatlands.     

An improved NDVI-based threshold method for estimating land surface emissivity using MODIS satellite data

This work estimated the land surface emissivities (LSEs) for MODIS thermal infrared channels 29 (8.4–8.7 μm), 31 (10.78–11.28 μm), and 32 (11.77–12.27 μm) using an improved normalized difference vegetation index (NDVI)-based threshold method. The channel LSEs are expressed as functions of atmospherically corrected reflectance from the MODIS visible and near-infrared channels with wavelengths ranging from 0.4 to 2.2 μm for bare soil. To retain the angular information, the vegetation LSEs were explicitly expressed in the NDVI function. The results exhibited a root mean square error (RMSE) among the estimated LSEs using the improved method, and those calculated using spectral data from Johns Hopkins University (JHU) are below 0.01 for channels 31 and 32. The MODIS land surface temperature/emissivity (LST/E) products, MOD11_L2 with LSE derived via the classification-based method with 1 km resolution and MOD11C1 with LSE retrieved via the day/night LST retrieval method at 0.05° resolution, were used to validate the proposed method. The resultant variances and entropies for the LSEs estimated using the proposed method were larger than those extracted from MOD11_L2, which indicates that the proposed method better described the spectral variation for different land covers. In addition, comparing the estimated LSEs to those from MOD11C1 yielded RMSEs of approximately 0.02 for the three channels; however, more than 70% of pixels exhibited LSE differences within 0.01 for channels 31 and 32, which indicates that the proposed method feasibly depicts LSE variation for different land covers.     

Quantifying scaling effects on satellite‐derived forest area estimates for the conterminous USA

We quantified the scaling effects on forest area estimates for the conterminous USA using regression analysis and the National Land Cover Dataset 30 m satellite‐derived maps in 2001 and 1992. The original data were aggregated to: (1) broad cover types (forest vs. non‐forest); and (2) coarser resolutions (1 km and 10 km). Standard errors of the model estimates were 2.3% and 4.9% at 1 km and 10 km resolutions, respectively. Our model improved the accuracies for 1 km by 0.6% (12 556 km²) in 2001 and 1.9% (43 198 km²) in 1992, compared to the forest estimates before the adjustments. Forest area observed from Moderate Resolution Imaging Spectroradiometer (MODIS) 2001 1 km land‐cover map for the conterminous USA might differ by 80 811 km² from what would be observed if MODIS was available at 30 m. Of this difference, 58% (46 870 km²) could be a relatively small net improvement, equivalent to 1444 Tg (or 1.5%) of total non‐soil forest CO₂ stocks. With increasing attention to accurate monitoring and evaluation of forest area changes for different regions of the globe, our results could facilitate the removal of bias from large‐scale estimates based on remote sensors with coarse resolutions.     

Remote sensing with IDRISI Taiga: a beginner's guide,by Timothy A. Warner and David J.

ABSTRACT

We applied annual Moderate Resolution Imaging Spectroradiometer (MODIS) product Vegetation Continuous Fields (VCF) data for the detection of forest cover change (FCC) in Mexico over the period 2000–2010. We excluded the pixels with uncertain information and applied a moving average and low-pass filter to smooth the multi-temporal data to reduce the fluctuations in the forest cover for each pixel. We applied a linear regression model and created two scenarios based on the coefficient of determination and slope to determine whether a pixel had changed its land cover over the study period. This model was able to label detected changes as deforestation, degradation, reforestation, and regrowth, based on the initial and final values of forest cover. The results showed that there has been more forest gain (reforestation and regrowth) than forest loss (deforestation and degradation) during the study period. We verified these results by comparing with the biomass data derived from the Mexican National Forest and Soil Inventory (Inventario Nacional Forestal y de Suelos, abbreviated to INFyS). Our model provides an efficient method to assess FCC at national level, which can contribute to the development of a reference level of greenhouse gas emission as necessary for the implementation of the international policy for reduction of emissions from deforestation and forest degradation (REDD+).     

Cropland distributions from temporal unmixing of MODIS data

Knowledge of the distribution of crop types is important for land management and trade decisions, and is needed to constrain remotely sensed estimates of variables, such as crop stress and productivity. The Moderate Resolution Imaging Spectroradiometer (MODIS) offers a unique combination of spectral, temporal, and spatial resolution compared to previous global sensors, making it a good candidate for large-scale crop type mapping. However, because of subpixel heterogeneity, the application of traditional hard classification approaches to MODIS data may result in significant errors in crop area estimation. We developed and tested a linear unmixing approach with MODIS that estimates subpixel fractions of crop area based on the temporal signature of reflectance throughout the growing season. In this method, termed probabilistic temporal unmixing (PTU), endmember sets were constructed using Landsat data to identify pure pixels, and uncertainty resulting from endmember variability was quantified using Monte Carlo simulation. This approach was evaluated using Landsat classification maps in two intensive agricultural regions, the Yaqui Valley (YV) of Mexico and the Southern Great Plains (SGP). Performance of the mixture model varied depending on the scale of comparison, with R² ranging from roughly 50% for estimating crop area within individual pixels to greater than 80% for crop cover within areas over 10 km². The results of this study demonstrate the importance of subpixel heterogeneity in cropland systems, and the potential of temporal unmixing to provide accurate and rapid assessments of land cover distributions using coarse resolution sensors, such as MODIS.