Leaf Area Index Distribution Research Articles

An analytical algorithm for the determination of vegetation Leaf Area Index from TRMM/TMI data

In this paper, an analytical algorithm for the determination of land surface vegetation Leaf Area Index (LAI) with the passive microwave remote sensing data is developed. With the developed algorithm and the Tropical Rainfall Measuring Mission/Microwave Imager (TRMM/TMI) remote sensing data collected during the Global Energy and Water Experiment (GEWEX) Asian Monsoon Experiment in Tibet (GAME/Tibet) Intensive Observation Period (IOP'98), the regional and temporal distributions of the land surface vegetation LAI have been evaluated. To validate the developed algorithm and the retrieval results, the maximum-composite Normalized Difference Vegetation Index (NDVI) data over the same study area and period are used in this study; the cloud contaminated NDVI values have been replaced by the cloud-free values reconstructed by the Harmonic ANnalysis of Time Series (HANTS) technique. The results show that the retrieved LAI is in good agreement with the cloud-free NDVI in regional and temporal distributions and in their statistical characteristics; the vegetation characteristics can be clearly assessed from the regional distribution of the retrieved LAI. As lower frequency microwave radiation can penetrate atmosphere and thin cloud layer, with the application of the passive microwave remote sensing data, the developed algorithm can be used to monitor the land surface vegetation condition more effectively.

Evaluating image-based estimates of leaf area index in boreal conifer stands over a range of scales using high-resolution CASI imagery

Leaf area index (LAI) is an important surface biophysical parameter as a measure of vegetation cover, vegetation productivity, and as an input to ecosystem process models. Recently, a number of coarse-scale (1-km) LAI maps have been generated over large regions including the Canadian boreal forest. This study focuses on the production of fine-scale (≤30-m) LAI maps using the forest light interaction model-clustering (FLIM-CLUS) algorithm over selected boreal conifer stands and the subsequent comparison of the fine-scale maps to coarse-scale LAI maps synthesized from Landsat TM imagery. The fine-scale estimates are validated using surface LAI measurements to give relative root mean square errors of under 7% for jack pine sites and under 14% for black spruce sites. In contrast, finer scale site mean LAI ranges between 49% and 86% of the mean of surface estimates covering only part of the sites and 54% to 110% of coarse-scale site mean LAI. Correlations between fine-scale and coarse-scale estimates range from near 0.5 for 30-m coarse-scale images to under 0.3 to 1-km coarse-scale images but increase to near 0.90 after imposing fine-scale zero LAI areas in coarse-scale estimates. The increase suggests that coarse-scale image-based LAI estimates require consideration of sub-pixel open areas. Both FLIM-CLUS and coarse-scale site mean LAI are substantially lower than surface estimates over northern sites. The assumption of spatially random residuals in regression-based estimates of LAI may not be valid and may therefore add to local bias errors in estimating LAI remotely. Differences between fine-scale airborne LAI maps and 30-m-scale Landsat TM LAI maps suggests that, for sparse boreal conifer stands, LAI maps produced from Landsat TM alone may not always be sufficient for validation of coarser scale LAI maps. In addition, previous studies may have used biased LAI estimates over the study site. Fine-scale spatial LAI maps offer one means of assessing and correcting for effects of sub-pixel open area patches and for characterising the spatial pattern of residuals in coarse-scale LAI estimates in comparison to the true distribution of LAI on the surface.

Impact of leaf area index seasonality on the annual land surface evaporation in a global circulation model

Numerical experiments with the global model of the European Centre for Medium‐Range Weather Forecasts (ECMWF) were devoted to the sensitivity of the modeled evaporation and precipitation to the vegetation leaf area index (LAI). The temporally static LAI distribution was replaced with a seasonally varying LAI, derived from normalized difference vegetation index (NDVI) archives. The seasonality of surface evaporation increased likewise, in combination with a response of increased precipitation seasonality over land. In a second set of experiments, the LAI estimates were perturbed by a noise term reflecting measurement accuracy and interannual variability. The resulting noise in evaporation and precipitation was compared to the noise intrinsically generated by the atmosphere. For periods and areas where evaporation forms a large term in the surface energy balance, the noise added to the LAI could be clearly discerned from the atmospheric noise, indicating that improved LAI estimation techniques can have a detectable impact on the surface evaporation calculated in the ECMWF global model.

Deciduous forest communities in the Black Sea coastal Strandzha region: temporal and spatial characteristics of leaf area index and density

The seasonal variation and vertical distribution of leaf area index are important characteristics of forest stands and their structure, not in the least for modelling and assessing biogeochemical and biogeophysical processes. In this paper results are presented from 2 years' measurements with the Li-Cor LAI-2000 Plant Canopy Analyser in the south-eastern part of the Bulgarian Black Sea coast, where Quercus frainetto stands are well represented but so far little studied. Mixed Fagus orientalis and Quercus polycarpa stands were also studied. The Q. frainetto effective LAI-2000 leaf area index values (one-sided) ranged from 1.43 to 3.21 with maximum values in June. Several Q. frainetto and F. orientalis + Q. polycarpa stands in the coastal zone were studied extensively and are shown to follow the seasonal variation at an intensive-measurements site, with an increase in leaf area index by 39.6% and 45.5% in June, respectively, related to the April values. The vertical profile of the canopy in the intensive-measurements site is presented and destructive measurements are compared with the LAI-2000 data. Data show that most of the leaves are located close to the middle of tree height (at around 40% of it). The analysis suggests that the intensive-measurements site is a good representative of the region, and that LAI-2000 measurements can provide sufficiently good information for further mapping of its forest leaf-area characteristics.

Integrating clumping effects in forest canopy structure: an assessment through hemispherical photographs

Methods for analysing foliage nonrandomness in forest canopies by means of hemispherical photographs are assessed. These methods involve calculation of the canopy element clumping factor, at scales coarser than that of the shoot, to adjust for clumping effects on leaf area index (LAI) estimates derived from gap fraction measurements. Two approaches are presented. The first is based on a gap size accumulation method (the Chen and Cihlar clumping index), whereas the second relies on a gap size distribution method (the Pielou coefficient of segregation). Both methods take advantage of hemispherical photographs, allowing measurement of gap size and gap fraction from sequences of black (foliage) and white (sky) pixels along circular transects over the whole range of angles. Clumping factors generated by hemispherical photographs have been analysed and compared using (i) hemispherical photographs of simulated forest stands with fixed LAI (2, 4, and 6) and three-dimensional spatial distribution of foliage, ranging from complete randomness to full clumping; and (ii) in situ hemispherical photographs from forest canopies with known architecture. The following conclusions are drawn based on our experimental data set: (i) overall clumping factors can be obtained by integrating values over the same range of angles as that used to derive LAI, therefore ensuring consistency between estimations of LAI and clumping factors; (ii) the Chen and Cihlar clumping index tends to underestimate clumping in highly clumped canopies, whereas the Pielou coefficient of segregation tends to overestimate clumping in poorly clumped canopies; and (iii) hemispherical photographs provide an efficient tool for describing the degree of canopy nonrandomness in all directions of the hemisphere and to adjust clumping effects on LAI estimates derived from gap fraction analysis. However, clumping factors derived from hemispherical photographs need to be further tested in real canopies and compared with other methods to define their merits and limitations.

Scaling the spatial distribution of photosynthesis from leaf to canopy in a plant growth chamber

The aim of this study was to quantify the spatial distribution of the net photosynthesis of a multi-layer crop. A multi-layer model (FORUG) was used to simulate the crop photosynthesis of tomato ( Lycopersicon esculentum Mill.) in the three-dimensional space inside a plant growth chamber. The model inputs were, on the one hand photosynthetic characteristics (maximal photosynthetic rate, quantum efficiency and respiration rate) determined at leaf level, and the microclimatic data on the other hand. Scaling-up to the level of the crop was done, taking into account the spatial distribution of leaf area index (LAI), leaf angle (or extinction coefficient), air temperature and photosynthetically active radiation (PAR). The simulated output was the net carbon exchange at a specific time step or the cumulative net carbon exchange over a specific period of time. Inside the test chamber spatial temperature differences ranged from 22.4 up to 32.9 °C at an inlet temperature of 22 °C and the lights on. Light intensity decreased gradually within the crop, resulting in very low light intensities at the bottom layers of the climate chamber. This resulted in the observation that the respiration rate was not compensated at low light intensities. The simulated photosynthesis of the 32 ‘cells’ of the test chamber ranged from −1.12 up to 5.94 μmol CO 2 m −2 s −1. From a sensitivity analysis of the model at low (20% of maximum) and high (80% of maximum) light intensities, it was concluded that the spatial distribution of light, air temperature and LAI should be well known, just as the parameters describing the light response of the photosynthesis process. Moreover, this spatial distribution should be accurately taken into account.

A model for seasonality and distribution of leaf area index of forests and its application to China

Abstract. We have constructed a phenological model of leaf area index (LAI) of forests based on biological principles of leaf growth. Field data of maximum LAI from 794 plots with mature or nearly mature stand ages over China were used to parameterize and calibrate the model. New measurements of maximum LAI from 16 natural forest sites were used to validate the simulated maximum LAI. The predictions of seasonal LAI patterns were compared with seasonal changes derived from the 1‐km satellite AVHRR‐NDVI data for nine undisturbed forest sites in eastern China. Then, we used the model to map maximum LAI values for forests in China.Model results indicated that the PhenLAI model generally predicted maximum LAI well for most forest types, even when maximum LAI is > 6. This suggests an ecological approach to the saturation problem in satellite detection of high forest LAI where the relationship between NDVI and LAI reaches an asymptote near a projected LAI value of 5 or 6. Furthermore, the predictions of seasonal LAI patterns in timing and dynamics were generally consistent with the satellite NDVI changes, except for monsoon forest and rain forest in south China where satellite detection of seasonal variation in leaf area is hardly possible. Compared with average projected LAI measurements of global forests from 809 field plots in literature data, our maximum LAI values were close to the global literature data for most of Chinese forests, but the average area‐weighted maximum LAI for all forests of China (6.68 ± 3.85) was higher than the global mean LAI of the 809 field plots (5.55 ± 4.14). We believe that forest LAI in China is commonly > 6, especially in tropical rainforest, subtropical evergreen broad‐leaved forest, temperate mixed forest, and boreal/alpine spruce‐fir forest where satellite detection of high LAI is hardly possible.

A model for seasonality and distribution of leaf area index of forests and its application to China

We have constructed a phenological model of leaf area index (LAI) of forests based on biological principles of leaf growth. Field data of maximum LAI from 794 plots with mature or nearly mature stand ages over China were used to parameterize and calibrate the model. New measurements of maximum LAI from 16 natural forest sites were used to vali- date the simulated maximum LAI. The predictions of seasonal LAI patterns were compared with seasonal changes derived from the 1-km satellite AVHRR-NDVI data for nine undis- turbed forest sites in eastern China. Then, we used the model to map maximum LAI values for forests in China. Model results indicated that the PhenLAI model generally predicted maximum LAI well for most forest types, even when maximum LAI is > 6. This suggests an ecological approach to the saturation problem in satellite detection of high forest LAI where the relationship between NDVI and LAI reaches an asymptote near a projected LAI value of 5 or 6. Furthermore, the predictions of seasonal LAI patterns in timing and dynam- ics were generally consistent with the satellite NDVI changes, except for monsoon forest and rain forest in south China where satellite detection of seasonal variation in leaf area is hardly possible. Compared with average projected LAI measure- ments of global forests from 809 field plots in literature data, our maximum LAI values were close to the global literature data for most of Chinese forests, but the average area-weighted maximum LAI for all forests of China (6.68 ± 3.85) was higher than the global mean LAI of the 809 field plots (5.55 ± 4.14). We believe that forest LAI in China is commonly > 6, espe- cially in tropical rainforest, subtropical evergreen broad-leaved forest, temperate mixed forest, and boreal/alpine spruce-fir forest where satellite detection of high LAI is hardly possible.

Season‐Long Characterization of Vertical Distribution of Leaf Area in Corn

Amount and vertical distribution of leaf area are essential for estimating radiation interception for canopy photosynthesis modeling. The objective of this study was to quantify vertical distribution of leaf area in corn (Zea mays L.) during the entire growing season. Field experiments were conducted in 1994 and 1995 near Mead, NE. Treatments included irrigation vs. no irrigation and three N rates (0, 68, and 135 kg ha−1). A leaf‐dissecting device was built for accurate clipping and collection of leaves at 0.10‐m height intervals. A normal curve gave an excellent fit to measured vertical distributions of leaf area index (LAI) at 0.10‐m height intervals (r2 > 0.94), regardless of applied N, soil moisture, crop developmental stage, or year. It was concluded that corn LAI was distributed symmetrically in the vertical from crop emergence to maturity. Plots of LAI vertical distributions indicated not only the onset and magnitude of differences among treatments, but also how stress was shared among all canopy layers. Of the three fitted curve parameters, height of symmetry and measure of green‐canopy spread were good indicators of plant growth, but LAI in interval containing height of the symmetry axis was not. The existence of a symmetrical vertical distribution of LAI fulfilled the requirement for applying the three‐point Gaussian method of integration in corn.

GROSS PRIMARY PRODUCTIVITY IN DUKE FOREST: MODELING SYNTHESIS OF CO2EXPERIMENT AND EDDY–FLUX DATA

This study was designed to estimate gross primary productivity (GPP) in the Duke Forest at both ambient and elevated CO2 (ambient + 200 fLL/L) concentrations using a physiologically based canopy model. The model stratified the canopy of loblolly pine (Pinus taeda L.) forest into six layers and estimated photosynthesis in each layer according to the Farquhar sub model coupled with the Ball-Berry stomatal conductance sub model. The model was parameterized with a suite of physiological measurements, in- cluding leaf area index (LAI), leaf nitrogen (N) concentration, photosynthesis-N relation- ships, and stomatal conductance. The model was validated against measured leaf photo- synthesis and canopy carbon (C) fluxes estimated from eddy-covariance measurements (ECM). Application of this model to simulate canopy C fixation from 28 August 1996, the onset of CO2 fumigation, to 31 December 1998 suggested that elevation of atmospheric (CO2J to ambient + 200 fLL/L resulted in increase of canopy C fixation by 35% in 1996, 39% in 1997, and 43% in 1998. The modeled GPP and its response to elevated (CO2J were sensitive to parameter values of quantum yield of electron transport, leaf area index, and the vertical distribution of LAI within the canopy. Thus, further investigation on those parameters will help improve the precision of estimated ecosystem-scale C fluxes. Fur- thermore, comparison between the modeled and ECM-estimated canopy C fluxes suggested that soil moisture, in addition to air vapor pressure, controlled canopy photosynthesis during the drought period.

Predicting spatial and temporal patterns of soil temperature based on topography, surface cover and air temperature

Soil temperature is a variable that links surface structure to soil processes and yet its spatial prediction across landscapes with variable surface structure is poorly understood. In this study, a hybrid soil temperature model was developed to predict daily spatial patterns of soil temperature in a forested landscape by incorporating the effects of topography, canopy and ground litter. The model is based on both heat transfer physics and empirical relationship between air and soil temperature, and uses input variables that are extracted from a digital elevation model (DEM), satellite imagery, and standard weather records. Model-predicted soil temperatures fitted well with data measured at 10 cm soil depth at three sites: two hardwood forests and a bare soil area. A sensitivity analysis showed that the model was highly sensitive to leaf area index (LAI) and air temperature. When the spatial pattern of soil temperature in a forested watershed was simulated by the model, different responses of bare and canopy-closed ground to air temperature were identified. Spatial distribution of daily air temperature was geostatistically interpolated from the data of weather stations adjacent to the simulated area. Spatial distribution of LAI was obtained from Landsat Thematic Mapper images. The hybrid model describes spatial variability of soil temperature across landscapes and different sensitivity to rising air temperature depending on site-specific surface structures, such as LAI and ground litter stores. In addition, the model may be beneficially incorporated into other ecosystem models requiring soil temperature as one of the input variables.

Leaf Area Index Estimates Using Remotely Sensed Data and BRDF Models in a Semiarid Region

The amount and spatial and temporal dynamics of vegetation are important information in environmental studies and agricultural practices. There has been a great deal of interest in estimating vegetation parameters and their spatial and temporal extent using remotely sensed imagery. There are primarily two approaches to estimating vegetation parameters such as leaf area index (LAI). The first one is associated with computation of spectral vegetation indices (SVI) from radiometric measurements. This approach uses an empirical or modeled LAI–SVI relation between remotely sensed variables such as SVI and biophysical variables such as LAI. The major limitation of this empirical approach is that there is no single LAI-SVI equation (with a set of coefficients) that can be applied to remote-sensing images of different surface types. The second approach involves using bidirectional reflectance distribution function (BRDF) models. It inverts a BRDF model with radiometric measurements to estimate LAI using an optimization procedure. Although this approach has a theoretical basis and is potentially applicable to varying surface types, its primary limitation is the lengthy computation time and difficulty of obtaining the required input parameters by the model. In this study, we present a strategy that combines BRDF models and conventional LAI–SVI approaches to circumvent these limitations. The proposed strategy was implemented in three sequential steps. In the first step, a BRDF model was inverted with a limited number of data points or pixels to produce a training data set consisting of leaf area index and associated pixel values. In the second step, the training data set passed through a quality control procedure to remove outliers from the inversion procedure. In the final step, the training data set was used either to fit an LAI–SVI equation or to train a neural fuzzy system. The best fit equation or the trained fuzzy system was then applied to large-scale remote-sensing imagery to map spatial LAI distribution. This approach was applied to Landsat TM imagery acquired in the semiarid southeast Arizona and AVHRR imagery over the Hapex-Sahel experimental sites near Niamy, Niger. The results were compared with limited ground-based LAI measurements and suggested that the proposed approach produced reasonable estimates of leaf area index over large areas in semiarid regions. This study was not intended to show accuracy improvement of LAI estimation from remotely sensed data. Rather, it provides an alternative that is simple and requires little knowledge of study target and few ground measurements.

Performance of the Canadian land surface scheme at temperate aspen‐birch and mixed forests, and a boreal young jack pine forest: Tests involving canopy conductance parametrizations

The Canadian Land Surface Scheme (CLASS) was evaluated using off‐line tests, with data collected at temperate aspen‐birch and mixed forests located near Chalk River, Ontario, and at a young jack pine forest located near Thompson, Manitoba. Measurements of stomatal conductance and leaf area index were used to develop canopy conductance parametrizations for the aspen‐birch and the jack pine stands, and these were incorporated into CLASS in a modified run. The behaviour of the residuals from the unmodified run with respect to environmental variables at the mixed stand were used as a guide to producing a modified ad‐hoc canopy conductance parametrization for use at that site. It was found that CLASS exaggerates the diurnal range in soil heat flux. This leads to an underestimation of the diurnal range in available energy, and it is argued that this has implications for its partitioning into the turbulent fluxes of sensible and latent heat. Over time, systematic errors will accumulate which, in the case of the latent heat flux, can affect the long‐term water balance. Errors in the size of the turbulent fluxes over short time periods can limit the usefulness of CLASS for resolving mesoscale circulations in a regional model. These problems are believed to be related to CLASS's assumption of zero wind speed under the canopy, and the lack of an insulating litter layer at the soil surface. An unmodified CLASS produced estimates of sensible and latent heat fluxes that broadly agreed with measured values at the temperate stands but not at the jack pine stand, where the latent heat flux was significantly overestimated. The modified canopy conductance parametri‐zations improved the representation of the latent heat flux at all of the sites, particularly at the jack pine stand. Overestimation at the aspen‐birch stand may be related to site heterogeneity not being adequately represented in the measurement of stomatal conductance, or to difficulties in defining the vertical distribution of leaf area index for scaling stomatal conductance to the canopy level. As expected, a positive or negative bias in the modelled latent heat flux is associated with an opposite bias in the modelled sensible heat flux. The tendency towards positive intercepts and slopes less than unity in linear regressions relating modelled and measured fluxes is a symptom of the underestimation of the diurnal range in available energy. Recommendations are made to allow for vertical discontinuities in texture within a soil layer and a variable permeable depth, and to allow transpiration when the soil water suction exceeds –1.5 MPa. This is particularly important when soil moisture or texture is heterogeneous. Finally, it is suggested that volumetric soil moistures should be allowed to fall below the originally prescribed limiting value of 0.04. This is necessary for evaporation to continue in very sandy, dry soils.

Modeling of the radiation regime and photosynthesis of a finite canopy using the DART model. Influence of canopy architecture assumptions and border effects

The scope of this work was to investigate the impact of the border effects and the 3-D architecture of a fallow field on: 1) its bidirectional reflectance factor (BRF); 2) its PAR (photosynthetically active radiation) regime; and, to a lesser extent, 3) on its carbon assimilation. For this purpose, laboratory BRF measurements were conducted on a sample of a fallow field. Moreover, we modified a 3-D radiative transfer model in order to simulate the visible and near infrared BRF of finite and heterogeneous media. Several scene representations were used (finite and infinite scenes with/without 1-D or 3-D distribution of leaf area index [LAI]). Results showed that border effects and LAI distribution strongly affect the BRF, with variations as large as 40% depending on the scene representations and on the spectral domain. PAR profiles and instantaneous canopy carbon assimilation of an infinite scene (natural conditions) were also studied with the 3-D model. The results stressed that, in the case of a fallow field, the use of a simple LAI profile provides enough information to accurately simulate the effects of the architecture on the PAR regime and the carbon assimilation of a fallow field.

Prototyping of MISR LAI and FPAR algorithm with POLDER data over Africa

The multi-angle imaging spectroradiometer (MISR) instrument is designed to provide global imagery at nine discrete viewing angles and four visible/near-infrared spectral bands. The MISR standard products include vegetation canopy green leaf area index (LAI) and fraction of photosynthetically active radiation absorbed by vegetation (FPAR). These products are produced using a peer-reviewed algorithm documented in the EOS-AM1 (Terra) special issue of the Journal of Geophysical Research. This paper presents results on spatial distributions of LAI and FPAR of vegetated land surfaces derived from the MISR LAI/FPAR algorithm with bidirectional reflectance data from the polarization and directionality of the Earth's reflectance (POLDER) instrument over Africa. The results indicate that the proposed algorithm reflects the physical relationships between surface reflectances and biophysical parameters and demonstrates the advantages of using multiangle data instead of single-angle data. A new method for evaluating bihemispherical reflectance (BHR) from multi-angle measurements of hemispherical directional reflectance factor (HDRF) was developed to prototype the algorithm with POLDER data. The accuracy of BHR evaluation and LAI/FPAR estimation is also presented. To authors demonstrate the advantages of using multi-angle data over single-angle data of surface reflectance.

Leaf area and tree increment dynamics of even-aged and multiaged lodgepole pine stands in Montana

Age structure and distribution of leaf area index (LAI) of even and multiaged lodgepole pine (Pinus contorta var. latifolia Engelm.) stands were examined on three study areas in western and central Montana. Projected leaf area was determined based on a relationship with sapwood cross-sectional area at breast height. Stand structure and LAI varied considerably between individual plots. LAI and stand stem volume increment were significantly higher in multiaged than even-aged stands with the exception of one study area, which had higher volume increment in even-aged stands. Older cohorts and higher canopy strata generally had greater LAI than younger cohorts and lower strata. Ratios of stem volume increment to leaf area were used to assess stand, cohort, and individual tree vigor or growing space efficiency (GSE). Even-aged stands had significantly higher GSEs in individual study areas and overall than multiaged stands. Cohort GSE generally increased with increasing age of the cohort. Stand increment was weakly associated with stand LAI. Individual tree volume increment was strongly related to projected leaf area when stands were divided by age-classes or canopy strata. These results suggest separating these stands into components, such as age classes or canopy strata, and summing predicted increment for each component may provide more accurate prediction of stand increment than using whole-stand LAI.

Measurements of leaf orientation, light distribution and sunlit leaf area in a boreal aspen forest

A new instrument called a Multiband Vegetation Imager (MVI) ( Kucharik et al., 1997), which uses a 16-bit charge-coupled device (CCD) camera and filter exchange mechanism to capture 2-band (visible and near-infrared) image pairs of plant canopies, has been used to measure the light distribution over sunlit leaves and indirectly infer leaf area index (LAI), sunlit LAI and leaf angle distribution (LAD) in a boreal aspen ( Populus tremuloides) forest during the Boreal Ecosystem–Atmosphere Study (BOREAS). One purpose of this study is to demonstrate that by combining MVI measurements with numerical Monte Carlo simulations of forest canopy architecture, the LAD and sunlit LAI of aspen can be obtained indirectly. Our results show that this boreal aspen stand exemplifies an erectophile LAD, with a mean leaf inclination angle near 70°. We also find that the values of the measured and modeled sunlit leaf area in aspen do not change dramatically for typical northern boreal latitude solar zenith angles (i.e. 30–70°). A major problem with determining the sunlit LAI is deciding what range of light intensities constitute a sunlit leaf because penumbra create a smooth continuum of light intensities over sunlit and shaded leaves in the canopy. Therefore, we show that the upper and lower limits can be placed on sunlit LAI values in aspen by using different leaf illumination threshold levels to determine sunlit and shaded LAI in the canopy. Typically, sunlit LAI values in aspen (LAI=3.3) range between 0.8–1.0 at a 70° sun zenith angle and 1.1–1.6 at a 30° sun zenith angle. Monte Carlo simulations and MVI measurements suggest that canopy sunlit leaf area estimates are possible from below the canopy at modest LAI values (<∼4.0) but become unreliable in higher LAI (>5.0). Since a substantial fraction of the total sunlit leaf area can be viewed from below the canopy in aspen (40–60% of the total sunlit LAI), a representative light distribution can be measured and used to quantify the canopy leaf angle distribution.

Three dimensional model of the interception of light by a canopy

We present a computer model that simulates, in three dimensions, the interception of light by a canopy. The model divides the canopy into cubic cells, each one characterized by mean leaf angle and the leaf area index. The model calculates the probability that a beam will penetrate to any given cell without being intercepted by the foliage in the path, using an exponential extinction function. Penetration of direct and diffuse light is calculated separately. The model also considers the reflection and transmission of light. The model has been tested against real measurements taken in an artificial plantation, reaching an r 2 of 0.714 between predicted and observed values. A sensitivity test of the model is also presented. The parameters analyzed are: total Leaf Area Index (LAI), vertical distribution of LAI, incident radiation, mean angle of leaves, transmission coefficient of leaves and elevation of the sun. A stepwise regression analysis of the results let us identify the variables which most affect the model outputs: Leaf Area Index, elevation of the sun, mean leaf angle, and incident radiation. The transmission coefficient of leaves does not exert significant influence on the intercepted radiation. The proportions of direct and diffuse components in incoming radiation have a significant influence on the results.

Estimating leaf area index of mangroves from satellite data

The relationship between the normalised difference vegetation index (NDVI) and leaf area index (LAI) was modelled for mangroves growing on the Caicos Bank, Turks and Caicos Islands. NDVI values were used to predict LAI with this model and a thematic map of LAI produced from satellite data for the whole Bank. Mangrove LAI ranged between 0.8 and 7.0, with a mean of 3.96. LAI data, estimated from in situ measurements of canopy transmittance for a set of sites independent of those used to derive the LAI/NDVI model, were used to test the accuracy of this image. Accuracy was defined as the proportion of accuracy sites at which the LAI value (as estimated from field measurements) lay within the 95% confidence interval for the predicted value of LAI. The accuracy of this map was high (88%) and the mean difference between predicted and measured LAI was low (13%). Remote sensing is thus demonstrated as a powerful tool for estimating the spatial distribution of LAI for whole mangrove ecosystems. This information can be obtained rapidly compared to alternative methods of measuring LAI and can minimise the logistical and practical difficulties of fieldwork in inaccessible mangrove areas.

Field estimates of green leaf area index using laser-induced chlorophyll fluorescence

Leaf area index (LAI) is an important determinant of canopy photosynthesis, evapotranspiration and competition among crop plants and weeds. Direct methods of measuring LAI are labor-intensive. Previous indirect methods can be inaccurate under some sky conditions and fail to distinguish between green and senescent leaves. Most indirect methods are also somewhat invasive, in that the instrument must be placed below the crop canopy. We have modified the inclined-point quadrat method to estimate LAI, by substituting a laser beam for the traditional metal probe. Laser-induced chlorophyll fluorescence is used to detect living leaves (or other organs containing chlorophyll), allowing the crop canopy to be scanned from above. Field tests with sudangrass, wheat, and maize found reasonably good correlations ( r 2 = 0.76 to 0.98) between LAI estimates from our prototype instrument and from direct harvest of leaves. None of the crops tested had significant dead leaf area. This new method can also provide greater detail on the spatial distribution of LAI (e.g., across crop rows) than previous methods. Because of the limited range at which laser-induced chlorophyll fluorescence can be detected, this approach probably would not be useful for research on forests or tree crops.