Alemu Gonsamo

Spectral Response Function Comparability Among 21 Satellite Sensors for Vegetation Monitoring

Global and regional vegetation assessment strategies often rely on the combined use of multisensor satellite data. Variations in spectral response function (SRF) which characterizes the sensitivity of each spectral band have been recognized as one of the most important sources of uncertainty for the use of multisensor data. This paper presents the SRF differences among 21 Earth observation satellite sensors and their cross-sensor corrections for red, near infrared (NIR), and shortwave infrared (SWIR) reflectances, and normalized difference vegetation index (NDVI) aimed at global vegetation monitoring. The training data set to derive the SRF cross-sensor correction coefficients were generated from the state-of-the-art radiative transfer models. The results indicate that reflectances and NDVI from different satellite sensors cannot be regarded as directly equivalent. Our approach includes a polynomial regression and spectral curve information generated from a training data set representing a wide dynamics of vegetation distributions to minimize land cover specific SRF cross-sensor correction coefficient variations. The absolute mean SRF caused differences were reduced from 33.9% (20.1%) to 9.4 % (6%) for red, from 3.2 % (8.9%) to 1% (1.1% ) for NIR, from 2.9% (3.6 %) to 1.9% (1.6%) for SWIR, and from 7.1 % (9%) to 1.8% (1.7% ) for NDVI, after applying the SRF cross-sensor correction coefficients on independent top of canopy (top of atmosphere) data for all-embraced-sensor comparisons. Variations in processing strategies, non spectral differences, and algorithm preferences among sensor systems and data streams hinder cross-sensor spectra and NDVI comparability and continuity. The SRF cross-sensor correction approach provided here, however, can be used for studies aiming at large-scale vegetation monitoring with acceptable accuracy.

Improved LAI Algorithm Implementation to MODIS Data by Incorporating Background, Topography, and Foliage Clumping Information

Leaf area index (LAI) is one of the essential biogeophysical variables related to terrestrial carbon and biogeochemical cycles. The University of Toronto (UofT) LAI product is developed in order to support the European Space Agency GLOBCARBON project for global and climate change assessments. The climate and global change communities have recently requested for a daily 250-m LAI product in order to improve the spatial and temporal patterns of carbon pools and fluxes knowledge. In light of these considerations, we carry out further improvements on the UofT LAI algorithm, including enhanced spatial resolution (250 m) by considering an improved land cover map, local topography, clumping index, and background reflectance variations in order to produce canopy LAI time series. Here, we present the methodological framework and an evaluation of 250-m UofTv2 LAI estimates in forest stands of the Canadian Carbon Program fluxnet sites. The LAI distributions over Canada and the comparison with ground measurements show an improved LAI estimates from the UofT v2 LAI algorithm as compared with the UofT v1 LAI algorithm. One of the key differences between v1 and v2 UofT LAI product is that the former produces total LAI whereas the latter produces overstorey LAI in forest and total LAI in other vegetated land cover types. A daily LAI product can further be extracted from the 10-day UofT v2 LAI time series by fitting various curve fitting algorithms. Although, we have shown the LAI product only over Canada, the algorithm can also be extended for a global 250-m LAI product.

Accelerating Forest Growth Enhancement due to Climate and Atmospheric Changes in British Colombia, Canada over 1956-2001

Changes in climate and atmospheric CO2 and nitrogen (N) over the last several decades have induced significant effects on forest carbon (C) cycling. However, contributions of individual factors are largely unknown because of the lack of long observational data and the undifferentiating between intrinsic factors and external forces in current ecosystem models. Using over four decades (1956–2001) of forest inventory data at 3432 permanent samples in maritime and boreal regions of British Columbia (B.C.), Canada, growth enhancements were reconstructed and partitioned into contributions of climate, CO2 and N after removal of age effects. We found that climate change contributed a particularly large amount (over 70%) of the accumulated growth enhancement, while the remaining was attributed to CO2 and N, respectively. We suggest that climate warming is contributing a widespread growth enhancement in B.C.s forests, but ecosystem models should consider CO2 and N fertilization effects to fully explain inventory-based observations.

Normalized sensitivity measures for leaf area index estimation using three-band spectral vegetation indices

A great number of spectral vegetation indices (SVIs) have been developed to estimate key biophysical parameters such as leaf area index (LAI). Considerable interest is often given to the local optimization, performance analysis and sensitivity of each spectral band and SVI for LAI estimation given that several confounding factors are present. In this regard, inclusion of shortwave infrared (SWIR) reflectance in traditionally near-infrared (NIR)-red (R)-based SVIs has played a great role for local optimization and increased sensitivity of SVIs to LAI. This study presents the enhanced and normalized sensitivity functions for evaluating (1) the sensitivity of each spectral band and SVI to LAI and (2) the generic performance analysis of empirical model to estimate LAI based on the SVIs. Several alternatives for three-band (NIR-R-SWIR) SVI modifications have been recommended and proven to be simplistic and unbiased way of local optimization.

Validating and Linking the GIMMS Leaf Area Index (LAI3g) with Environmental Controls in Tropical Africa

The recent Global Inventory Modeling and Mapping Studies (GIMMS) LAI3g product provides a 30-year global times-series of remotely sensed leaf area index (LAI), an essential variable in models of ecosystem process and productivity. In this study, we use a new dataset of field-based LAITrue to indirectly validate the GIMMS LAI3g product, LAIavhrr, in East Africa, comparing the distribution properties of LAIavhrr across biomes and environmental gradients with those properties derived for LAITrue. We show that the increase in LAI with vegetation height in natural biomes is captured by both LAIavhrr and LAITrue, but that LAIavhrr overestimates LAI for all biomes except shrubland and cropland. Non-linear responses of LAI to precipitation and moisture indices, whereby leaf area peaks at intermediate values and declines thereafter, are apparent in both LAITrue and LAIavhrr, although LAITrue reaches its maximum at lower values of the respective environmental

Evaluation of the GLC2000 and NALC2005 land cover products for LAI retrieval over Canada

Land cover information is an important input parameter for retrieving key land surface biophysical parameters, such as leaf area index (LAI), often parameterized with geometrical-optical properties distinctive among land cover types. This paper presents a comparative assessment and evaluation of the 1 km Global Land Cover (GLC2000) and the 250 m North American Land Cover (NALC2005) over Canada. We used a 30 m Circa 2000 Land Cover from agricultural regions of Canada as a reference dataset. The comparative assessment and evaluation were made at six generalized class levels that were categorized based on relevance for parameterizations of key land surface biophysical parameter retrieval algorithms. The overall per-pixel agreement between the GLC2000 and the NALC2005 was 63.4%. The overall accuracies using the Circa 2000 reference data were 62.3% and 65.5% for the GLC2000 and NALC2005 datasets, respectively. Based on the improved version 2 University of Toronto LAI algorithm, up to a 42% difference in LAI estimation was noted over Canada due to differences in the two regional land cover datasets. This study assessed the performance of the newly produced NALC2005 product and presents, for the first time, the often overlooked land cover characterization impact on large scale LAI estimation.

Citizen Science: linking the recent rapid advances of plant flowering in Canada with climate variability.

The timing of crucial events in plant life cycles is shifting in response to climate change. We use phenology records from PlantWatch Canada ‘Citizen Science’ networks to study recent rapid shifts of flowering phenology and its relationship with climate. The average first flower bloom day of 19 Canadian plant species has advanced by about 9 days during 2001–2012. 73% of the rapid and unprecedented first bloom day advances are explained by changes in mean annual national temperature, allowing the reconstruction of historic flower phenology records starting from 1948. The overall trends show that plant flowering in Canada is advancing by about 9 days per °C. This analysis reveals the strongest biological signal yet of climate warming in Canada. This finding has broad implications for niche differentiation among coexisting species, competitive interactions between species, and the asynchrony between plants and the organisms they interact with.

Citizen science: best practices to remove observer bias in trend analysis

Citizen science, time series records over long periods of time, and wide geographic areas offer many opportunities for scientists to answer questions that would otherwise be impractical to investigate. Citizen scientists currently play active roles in a wide range of ecological projects; however, observer biases such as varying perception of events or objects being observed and quality of observations present challenges to successfully derive interannual variability and trend statistics from time series records. It is recommended that citizen science records, particularly those involving events such as plant phenology, should not be directly averaged across sites. The interannual variability expressed as an anomaly and trend expressed as a regression slope should be calculated for each site. Only the site level anomaly and regression slopes should be averaged to suppress observer biases.

Large-scale leaf area index inversion algorithms from high-resolution airborne imagery

Large-scale leaf area index (LAI) inversion algorithms were developed to determine the LAI of a forest located in Gatineau Park, Canada, using high-resolution colour and colour infrared (CIR) digital airborne imagery. The algorithms are parameter-independent and developed based on the principles of optical field instruments for gap fraction measurements. Cloud-free colour and CIR images were acquired on 21 August 2007 with 35 and 60 cm nominal ground pixel size, respectively. Normalized Difference Vegetation Index (NDVI), maximum likelihood and object-oriented classifications, and principal component analysis (PCA) methods were applied to calculate the mono-directional gap fraction. Subsequently, LAI was derived from inversion and compared with ground measurements made in 54 plots of 20 by 20 m using hemispherical photography between 10 and 20 August 2007. There was high inter-correlation (the Pearson correlation coefficient, R > 0.5, p < 0.01) among LAI values inverted using the classifications and PCA methods, but neither were highly correlated with LAI inverted from the NDVI method. LAI inverted from the NDVI-based gap fraction significantly correlated with ground-measured LAI (R = 0.63, root mean square error (RMSE) = 0.52), while LAI inverted from the classification and PCA-derived gap fraction showed poor correlation with ground-measured LAI. Consequently, the NDVI method was used to invert LAI for the whole study area and produce a 20‐m resolution LAI map.

Winter teleconnections can predict the ensuing summer European crop productivity

Climate change impact assessment on agricultural crop productivity is becoming an important research arena given the increasing yield losses due to the high frequency of droughts in recent years and the anticipated prevalence of extreme events in future climate scenarios (1–3). It is said that by the middle of the 21st century, climate change probably will result in more frequent wheat crop failures across Europe (1). Most studies examine crop yield and physiology responses to externally forced climate change, particularly variations in temperature, precipitation, or atmospheric CO2 concentration. The impacts of internal climatic oscillations on crop productivity are often overlooked. Projections of climate change impacts on crop yields are therefore inherently uncertain (2, 3). However, farmers expect the quantification of the impacts and forecasting ability ahead of cropping season to operationally plan potential management options and produce an economically viable yield.