Research Papers

Analysis of vegetation dynamics and climatic variability impacts on greenness across Canada using remotely sensed data from 2000 to 2009

[+] Author Affiliations
Xiuqin Fang

Hohai University, School of Earth Sciences and Engineering, Nanjing 210098, China

University of Quebec at Montreal, Institute of Environment Sciences, Department of Biology Sciences, Montreal, QC, H3C 3P8 Canada

Qiuan Zhu

Northwest Agriculture and Forest University, College of Forestry, Laboratory for Ecological Forecasting and Global Change, Yanglin 712100, China

Huai Chen

Northwest Agriculture and Forest University, College of Forestry, Laboratory for Ecological Forecasting and Global Change, Yanglin 712100, China

Zhihai Ma

University of Quebec at Montreal, Institute of Environment Sciences, Department of Biology Sciences, Montreal, QC, H3C 3P8 Canada

Weifeng Wang

University of Quebec at Montreal, Institute of Environment Sciences, Department of Biology Sciences, Montreal, QC, H3C 3P8 Canada

McGill University, Department of Geography, 805 Sherbrooke Street W., Montreal, QC, H3A 0B9 Canada

Xinzhang Song

University of Quebec at Montreal, Institute of Environment Sciences, Department of Biology Sciences, Montreal, QC, H3C 3P8 Canada

Zhejiang Agriculture and Forestry University, State Key Laboratory of Subtropical Forest Science and Zhejiang Provincial Key Laboratory of Carbon Cycling and Carbon Sequestration in Forest Ecosystems, Lin’an 311300, China

Pengxiang Zhao

Northwest Agriculture and Forest University, College of Forestry, Laboratory for Ecological Forecasting and Global Change, Yanglin 712100, China

Changhui Peng

University of Quebec at Montreal, Institute of Environment Sciences, Department of Biology Sciences, Montreal, QC, H3C 3P8 Canada

Northwest Agriculture and Forest University, College of Forestry, Laboratory for Ecological Forecasting and Global Change, Yanglin 712100, China

J. Appl. Remote Sens. 8(1), 083666 (Mar 17, 2014). doi:10.1117/1.JRS.8.083666
History: Received July 6, 2013; Revised February 8, 2014; Accepted February 11, 2014
Text Size: A A A

Open Access Open Access

Abstract.  Using time series of moderate-resolution imaging spectroradiometer (MODIS) normalized difference vegetation index (NDVI) data from 2000 to 2009, we assessed decadal vegetation dynamics across Canada and examined the relationship between NDVI and climatic variables (precipitation and temperature). The Palmer drought severity index and vapor pressure difference (VPD) were used to relate the vegetation changes to the climate, especially in cases of drought. Results indicated that MODIS NDVI measurements provided a dynamic picture of interannual variation in Canadian vegetation patterns. Greenness declined in 2000, 2002, and 2009 and increased in 2005, 2006, and 2008. Vegetation dynamics varied across regions during the period. Most forest land shows little change, while vegetation in the ecozone of Pacific Maritime, Prairies, and Taiga Shield shows more dynamics than in the others. Significant correlations were found between NDVI and the climatic variables. The variation of NDVI resulting from climatic variability was more highly correlated to temperature than to precipitation in most ecozones. Vegetation grows better with higher precipitation and temperature in almost all ecozones. However, vegetation grows worse under higher temperature in the Prairies ecozone. The annual changes in NDVI corresponded well with the change in VPD in most ecozones.

Figures in this Article

Terrestrial ecosystems are continuously changing at a variety of spatial and temporal scales due to natural and/or anthropogenic causes.1 Understanding vegetation dynamics and how they respond to climate change is a fundamental requirement for predicting future ecosystem dynamics. Satellite remote sensing is an effective tool for monitoring vegetation at the regional and global scales.2 It provides full coverage of large and remote areas on a regular basis over extended periods of time as well as spatial data and associated information to deepen our understanding of ecological systems.3,4 Of the many remote sensing techniques for analyzing vegetation dynamics, time-series analysis of vegetation indices has become the most common approach for phenology and drought assessment.2 Among the numerous vegetation indices, the normalized difference vegetation index (NDVI), which is defined as the difference between near-infrared and red reflectances divided by their sum,5 is most commonly used to measure vegetation changes. Recently, moderate-resolution imaging spectroradiometer (MODIS) NDVI data have been widely used in vegetation studies and related drought monitoring. Studies that have applied time series of MODIS vegetation index data to explore forest disturbance and vegetation phenology, etc., throughout the world have demonstrated the applicability and feasibility of MODIS NDVI. Among others, we can mention studies on the Amazonian canopy,2 the continental United States,610 Alaska,11 Finland,12 Norway,13 Sweden,14 Australia,15 and northern Fennoscandia.16

Although remotely sensed data are the most widely used and effective means of studying global vegetation change, especially in light of climate change concerns,17 the conclusions on vegetation growth and the influence of climate changes differ across regions. For example, in terms of vegetation phenology, studies have demonstrated that vegetation growth in the mid-to-high latitudes of the northern hemisphere is very sensitive to temperature changes, particularly in spring.1821 A study in North America found a spring vegetation greening trend in response to higher spring temperatures in the northwestern region until the late 1980s or early 1990s, with an opposite trend in the northeastern region due to different spring temperature trends.22 However, a recent study on the Tibetan Plateau showed that winter and spring warming resulted in delayed spring phenology,23 whereas a more recent study implied that this delay could be attributed to other environmental factors.24 Meanwhile, the conclusions on the response of vegetation dynamics to a warming climate or drought vary across vegetation species and regions. In a study of boreal North America, the NDVI of evergreen coniferous forests revealed no consistent trend with temperature changes, whereas lengthier growing seasons were detected in the nearby tundra region.25 An analysis of Sahelian vegetation dynamics showed that vegetation condition patterns correlate with region-wide rainfall trends in the Sahel.26 A recent study revealed widespread reductions in the greenness of Amazon forests caused by the previous year’s record-breaking drought: NDVI maps showed that the 2010 drought reduced greenness over an approximately 2.5millionkm2 area in the Amazon—four times more than the area affected by the last severe drought in 2005.27

In Canada, ecosystem disturbances caused by insect infestations, diseases, windfall, and so on have been frequently studied using remotely sensed data in western Canada. As the mountain pine beetle epidemic in British Columbia (BC) represents a critical management and ecological concern, a lot of effort was made to detect red attack damage due to mountain pine beetle infestation using high-spatial resolution QuickBird imagery,28,29 IKONOS imagery,30 and aerial imagery31 or multidate Landsat imagery3234 and to assess changes in forest fragmentation35 or forest mortality using time-series Landsat imagery.36 Coops and Wulder37 estimated the reduction of gross primary production in BC due to the infestation between 2002 and 2005 using MODIS data and asserted that continued monitoring using MODIS-based approaches may offer ongoing opportunities to estimate the landscape-level rates of recovery from the outbreak. Using BC and Alberta as example areas, a disturbance index based on MODIS vegetation and temperature data was proposed and used to monitor land dynamics. The results indicated that the MODIS-based index is a useful tool to aid in change detection and national monitoring activities.38 Hilker et al.39 proposed a new data fusion algorithm for high spatial- and temporal-resolution mappings of forest disturbance based on Landsat and MODIS and tested it over a 185×185 km study area in west-central Alberta. Wulder et al.40 detected change events which occurred between 2000 and 2005 in Boreal Shield Ecozone in Ontario and asserted that MODIS, in combination with other remotely sensed data sources, can provide information on disturbance events within a national forest inventory remeasurement cycle. Fraser and Latifovic41 mapped defoliation and mortality using SPOT VEGETATION data over a coniferous forest region in Quebec that was severely defoliated during 1998 to 2000 by the eastern hemlock looper. Though frequent, numerous, and valuable vegetation dynamic studies in western Canada using remotely sensed data existed, there are fewer studies in the remaining regions in Canada. Especially, few cross-Canada studies have been conducted so far. The Crop Condition Assessment Program (CCAP) of the Statistics Canada Agency provides satellite NDVI information from 1987 for the whole Canadian agricultural area (http://www26.statcan.ca/ccap-peec/start-debut-eng.jsp). The Drought Watch Program of the Agriculture and Agri-Food Canada Agency aims to provide timely information of the impacts of climatic variability on water supply and agriculture from 2006 (http://www4.agr.gc.ca/AAFC-AAC/display-afficher.do?id=1256658312655&lang=eng). These are two valuable and practical long-term Canadian government programs. However, they are both limited to crop and pasture/rangeland conditions in Canada and reported on drought conditions and NDVI separately. Forests, as a big part of Canadian vegetation, are ignored in those programs. Consequently, crucial questions concerning Canadian vegetation dynamics and their response to climatic variability over the past decade remain to be answered. The objectives of this study were to: (1) analyze vegetation dynamics across Canada based on time series of NDVI data generated from Terra MODIS imagery over a 10-year period (2000 to 2009); and (2) assess the relationship between changing climate and vegetation greenness across different ecozones in Canada.

We used 16-day composite MODIS NDVI images at 1-km spatial resolution (MOD13A2, collection v005) to analyze seasonal and interannual vegetation dynamics and trends across Canada. Clouds influence the quality of remotely sensed data, which may cause NDVI data bias. We used the quality assessment (QA) layers to determine the validity of 16-day composite NDVI value. Specifically, only pixels with the following quality flags were considered valid: “MODLAND_QA” must be equal to 0 (good quality) or 1 (check other QA). “VI Usefulness” must be equal to 11 or less. “Adjacent cloud detected,” “Mixed clouds,” and “Possible shadow” flag values must be equal to 0. This QA method relating to cloud contamination has been used by Samanta et al.42

Eleven tiles of MODIS NVDI images (h09v04, h10v03, h10v04, h11v03, h11v04, h12v03, h12v04, h13v03, h13v04, h14v03, and h14v04) were acquired from February 2000 to December 2009 using the online search and order tool Global Visualization Viewer (GloVis; http://glovis.usgs.gov/) from U.S. Geological Survey (USGS). The selected tiles covered the region of 40° to 60°N, 50° to 140°W, for an overall area of 6.65millionkm2, or approximately 65% of the Canadian land mass (Fig. 1). All the Canadian terrestrial ecozones except Taiga Cordillera lie within the selected area (http://ecozones.ca/english/index.html). The most widespread vegetation type in the study area is forest. Canada’s National Forest Inventory 2006 (https://nfi.nfis.org/standardreports.php?lang=en#terrestrial) reported that above half of the ecozones areas are forest land in Taiga Plains, Boreal Shield, Atlantic Maritime, Boreal Plains, Pacific Maritime, and Montane Cordillera, especially in the ecozones of Boreal Shield and Atlantic Maritime with the forest land percentages of 80.0% and 83.2%, respectively. The above three ecozones occupy 74.8% forest land of the total study area. As the biggest ecozone in Canada, Boreal Shield is 29.9% of the study area and possesses 37.7% of the total forest land in Canada. The ecozones of Prairies and Mixedwood Plains possess 74.4% of the agricultural land of the total study area.

Graphic Jump LocationF1 :

Study area in Canadian territory.

We used a monthly reanalyzed 0.5°C gridded climate dataset for Canada from 2000 to 2009 obtained from the National Centre for Environmental Prediction (NCEP). The dataset contains grids of monthly temperature and precipitation for the Canadian landmass south of 60°N. The Palmer drought severity index (PDSI) and vapor pressure difference (VPD) at 0.5-deg resolution were used as drought indices to measure environmental water stress. PDSI was used as a surrogate for soil moisture by combining information from both evaporation and precipitation. Lower PDSI generally implies a drier climate.43 VPD is the difference between saturation vapor pressure (SVP) and air vapor pressure, and SVP is a nonlinear function of air temperature (T). As VPD is a major controlling factor for surface evapotranspiration (ET), high T increases both VPD and subsequent ET and leads to a drier environment.43

We focused on spatial patterns of annual mean NDVI and spatiotemporal patterns of NDVI during the growing season from 2000 to 2009. In a 16-day NDVI time series for each pixel, all of the 23 observations each year were used to calculate the annual mean NDVI. The months of April through October were selected to represent the average start and end of the growing season, similarly to Wang et al.,22 even though the start of the growing season may vary across the region from year to year. That is to say, 14 observations from April to October each year were used to represent NDVI values during the growing seasons for each pixel. Then, NDVI images for the growing season were averaged together to create one image that captured the greenness for each pixel. This may have slightly de-emphasized some of the greenest pixels. However, for a large area with a large number of pixels, it is the best way to obtain an image of what occurs during multiple growing seasons per year.44 Each growing season-averaged image from 2000 to 2009 was then averaged together to obtain a long-term average growing season NDVI for the decade. Year-to-year variability in NDVI patterns was determined by calculating annual growing season NDVI anomalies as follows: Display Formula

NDVIa=(NDVIgNDVIm)/NDVIm×100,(1)
where NDVIa is the growing season percent anomaly, NDVIg is the individual growing season averaged NDVI, and NDVIm is the decadal average growing season NDVI.

Correlation of Annual Mean NDVI with Climatic Variables

Figures 2(a), 2(b), and 2(c) show the annual mean distribution of NDVI, temperature, and precipitation, respectively. Figure 2(a) shows that in eastern parts of the study area, NDVI decreased with increasing latitude. In western parts, NDVI patterns are complex associated with topography. Figure 2(b) shows that the temperature generally decreased with increasing latitude. Figure 2(c) shows that the precipitation generally decreased with increasing latitude except for the western coastal region. The western coastal region and eastern regions show the highest annual precipitation.

Graphic Jump LocationF2 :

Average annual normalized difference vegetation index (NDVI) (a) derived from MODIS 16-day NDVI products, temperature (b), and precipitation (c) derived from National Centre for Environmental Prediction (NCEP).

Regression analyses were performed to quantify the relationships between annual mean NDVI and the two climatic variables (temperature and precipitation) using all corresponding vegetated pixels from Fig. 2. Regression analyses were carried out under two conditions with different target regions: the overall vegetated area and 13 ecozones regions with ecological variations (the Northern Arctic ecozone was excluded because it was nonvegetated in our study area). The regression analysis results are provided in Table 1.

Table Grahic Jump Location
Table 1Correlations between average annual NDVI and climatic variables with different target areas; + slope indicates a positive relationship and − slope indicates a negative relationship.

Table 1 shows that the correlations between average annual NDVI and the two climatic variables vary across the regions. For the overall vegetated area, we observe (1) a higher correlation between annual NDVI and temperature than between annual NDVI and precipitation; and (2) mostly positive slopes, indicating that the annual NDVI increased with increasing temperature and precipitation. For regions with ecological variation, distinctly higher correlations were observed between annual NDVI and precipitation in the ecozones of Arctic Cordillera, Hudson Plains, Prairies, and Taiga Plains than in the other ecozones. For these four ecozones, without exception, NDVI increased with increasing precipitation. Distinctly higher correlation between annual NDVI and temperature appeared in the ecozones of Arctic Cordillera, Boreal Shield, Hudson Plains, Prairies, Pacific Maritime, and Taiga Shield than in the other ecozones. For these six ecozones, NDVI increased with increasing temperature, except Prairies.

Furthermore, the correlation between NDVI and temperature was generally higher than between NDVI and precipitation. In addition, almost all the correlation slopes for the relationships between NDVI and precipitation are positive, which means that NDVI will increase with increasing precipitation. However, it was not to be the case for the slopes for the relationships between NDVI and temperature. NDVI increased with increasing temperature in some ecozones such as in Hudson Plains, but decreased with increasing temperature in other ecozones such as in Prairies.

Spatiotemporal NDVI and Drought Indices Patterns from 2000 to 2009
Changes in spatiotemporal NDVI patterns

The spatial NDVI anomaly patterns for the study area are shown in Fig. 3. These series of maps show the percent NDVI anomaly patterns for each growing season during the period (2000 to 2009).

Graphic Jump LocationF3 :

NDVI anomalies in growing seasons from 2000 to 2009.

In 2000 [Fig. 3(a)], three ecozones in western Canada (Pacific Maritime, Boreal Cordillera, and Montane Cordillera) and Taiga Shield and a small part of Prairies showed below-normal vegetation conditions, with most negative departures less than 15%. In contrast with 2000, most areas with below-normal vegetation conditions were reduced and showed greenness in 2001 [Fig. 3(b)]. However, the areas with below-normal vegetation conditions in the ecozone of Prairies extended from a small part of southern Alberta in 2000 to the whole Prairies ecozone in 2001 [Figs. 1 and 3(b)], which meant a prevalence of drought conditions over the southern Prairies provinces in 2001. In 2002 [Fig. 3(c)], most of the regions in the overall study area including the Prairies showed below-normal vegetation conditions or drought conditions. The lowest anomalies (<35%) were in western coastal Canada, especially in the ecozone of Montane Cordillera. However, situations changed in 2003 [Fig. 3(d)] as most of the regions showed normal vegetation conditions with anomalies between 5% and 5%. Particularly, the ecozone of Montane Cordillera in western coastal Canada showed pronounced greenness. In 2004 [Fig. 3(e)], the western coastal region showed more greenness than in 2003. Moreover, the southern Prairies provinces showed slight greenness, which is different from the above 4 years, with some positive departures above 15%. Taiga Shield ecozone in the northern Prairies provinces, northern boundary of Boreal Shield and Hudson Bay ecozone in central Canada, and some parts of Taiga Shield ecozone in eastern Canada showed below-normal vegetation conditions.

In 2005 and 2006 [Figs. 3(f) and 3(g)], most of the country showed above-normal vegetation conditions overall, indicating widespread greenness. In contrast, the western coastal region in 2007 [Fig. 3(h)] clearly showed below-normal vegetation conditions, with some departures at more than 35% in the ecozone of Pacific Maritime. In 2008 [Fig. 3(i)], the western coastal region and small parts of the northern Prairies provinces still showed slightly below-normal vegetation conditions, whereas northern parts of eastern Canada (Quebec and Newfoundland) showed substantial greenness. In 2009 [Fig. 3(j)], most of the overall area showed below-normal vegetation conditions, especially the Prairies provinces and eastern Canada, except for a small part of the western coastal region with slightly above-normal conditions. The vegetation situations in 2009 changed greatly compared with those in the last few years, especially showing patterns largely opposite of the ones in 2005 and 2006.

Generally, these series of NDVI anomalies show spatiotemporal vegetation dynamics and changes more marked throughout 2000s. The anomalies in 2000, 2002, and 2009 for most of the overall area are negative, indicating region-wide drought conditions. In contrast, the anomalies in 2005 and 2006 over most of the overall area are positive, indicating a region-wide above normal greening. Most forest land, such as the ecozones of Boreal Shield and Boreal Plains, shows relatively little change during the period of 2000 to 2009. Vegetation in the ecozone of Pacific Maritime, Prairies, and Taiga Shield shows more dynamics than in the other ecozones. The most negative anomalies in Pacific Maritime, Prairies, and Taiga Shield are in 2007, 2001, and 2000, respectively.

According to the USGS and U.S. Agency for International Development (http://earlywarning.usgs.gov/fews/index.php), NDVIa values between 5% and +5% indicate average vegetation conditions. NDVIa values below 5% represent below-average conditions, while those above +5% represent above-average conditions. Thus, in our study, a negative departure from normal NDVI value lower than 5% was considered drought conditions for vegetation growth, and a positive departure from normal NDVI value higher than +5% was considered favorable conditions for vegetation growth. Based on the above criteria, area statistics were used for 10 years for further quantitative comparisons. Table 2 presents all the area percentages for drought conditions and favorable conditions for each year. The three largest area percentages for drought conditions were approximately 44%, 37%, and 29% in 2002, 2009, and 2000, respectively, with the widest prevalence of drought conditions for vegetation growth, particularly in 2002. The three largest area percentages for the favorable conditions were about 45%, 36%, and 23% in 2005, 2006, and 2001, respectively, with the widest prevalence of favorable conditions for vegetation growth, particularly in 2005.

Table Grahic Jump Location
Table 2Area percentages of drought conditions and favorable conditions (%).
Analysis of spatiotemporal NDVI and drought indices patterns

Figure 3 and the above descriptions indicate that the decadal vegetation dynamics in the study area varied across regions from 2000 to 2009. Two drought indices (PDSI and VPD), which have been adapted for measuring environmental water stress,43 were used to relate the changes in the Canadian vegetation to the climate, especially in the cases of drought. The relation of both indices with drought is different. Generally, the PDSI is low while the VPD is high when drought is severe. Because of the unmatched spatial resolutions between NDVI anomalies and drought indices, which are at 1-km and 0.5-deg resolution, respectively, averages over each ecozone were analyzed to further explore spatiotemporal vegetation dynamics and drought conditions. Figure 4 shows the annual changes in NDVI, PDSI, and VPD averaged within each of the 13 ecozones.

Graphic Jump LocationF4 :

Annual changes in NDVI, PDSI, and VPD averaged within each of the 13 ecozones.

Figure 4 illustrates that the interannual variations of NDVI anomalies were unobvious in the ecozones of Atlantic Maritime, Boreal Plains, Mixedwoods Plains, and Taiga Plains, since the final shapes of the 2000 to 2009 change curves of NDVI anomalies are relatively smooth. The interannual variations of NDVI anomalies in the other ecozones were distinct but with respective features. For example, NDVI anomalies in the ecozones of Boreal Shield and Hudson Plains had similar variations and were relatively low in 2002, 2004, and 2009, whereas the relatively low NDVI anomalies in the ecozone of Prairies were observed in 2001 and 2002. Figure 4 also illustrates that the NDVI anomalies changed remarkably in the ecozone of Arctic Cordillera, where the lowest value appeared in 2008. In accordance with the conclusion at Sec. 3.2.1, we can also find from Fig. 4 that in most ecozones, vegetation conditions were below-normal ones commonly in 2002 and 2009 and above-normal ones commonly in 2005 and 2006.

With respect to the drought indices, we can find from Fig. 4 that both drought indices (PDSI and VPD) do not have the exactly same timing and indications of environmental drought conditions. For example, for Boreal Plains, PSDI indicated that the driest year was 2002 while VPD indicated that it was 2001. Furthermore, the timing of drought varies with ecozones regardless of considering any drought index. For example, drought indices indicated that the driest year in the ecozone of Prairies was 2001 while in Boreal Plains the driest one was 2002.

Comparing the interannual change patterns of NDVI anomalies and drought indices (Fig. 4), we can find out the NDVI anomaly curve for the ecozone of Prairies corresponds well with the PDSI curve but is converse to the VPD pattern. For the ecozone of Arctic Cordillera, NDVI anomaly curve was converse to the variations in VPD curve but corresponded relatively well with the PDSI curve from 2002 to 2009. However, for the remaining ecozones, the NDVI anomaly curves correspond better with VPD than that with PDSI.

Previous studies45,46 have suggested that some parts of Canada experienced one of the most serious and extensive droughts on record in 2001 and that the most severely affected areas were in the Prairies provinces, where the 2001 drought followed 2 to 3 consecutive years of below-average rainfall. Some parts of central Canada suffered the driest August on record, and western Canada underwent a second consecutive year of the most severe drought on record in 2002.47 In the present study, the vegetation dynamics in the study area differed across regions from 2000 to 2009. As presented above, the anomalies in 2000, 2002, and 2009 over most of the overall area are negative, indicating region-wide drought conditions and a clearly discernible area for the prevalence of drought conditions in 2002. The analyzed annual changes in NDVI averaged over ecologically different regions further confirmed the region-wide drought in 2002. The results of this study are consistent with recent related studies. For example, Hogg et al.48 examined the impacts of the 2001 to 2002 drought on aspen forests in western Canada and concluded that the drought led to a 30% decrease in regional aspen growth and was the main cause of aspen dieback and mortality. More recently, Peng et al.49 have reported that the water stress created by regional drought has contributed to the widespread increases in boreal tree mortality rates across Canada, but western Canada appears to have been more sensitive to drought than eastern Canada. Zhao et al.43 also found that droughts reduced net primary production (NPP) in North America in 2002. Recent reports by the Canadian Drought Research Initiative stated that the 2001 and 2002 droughts in Canada covered massive areas and were part of an unusually large, severe, and lengthy dry period.47,50

With respect to climatic variability impacts, significant correlations were found between NDVI values and the climatic variables (precipitation and temperature). Specifically, the correlations between NDVI and climatic variables varied with ecozones. With regard to spatial variations, higher correlations were found between NDVI and temperature than between NDVI and precipitation for most ecozones. However, in the ecozone of Taiga Plains, the correlation between NDVI and temperature was much lower than that between NDVI and precipitation. Almost all correlations between NDVI and climatic variables were positive, suggesting that vegetation grows better with higher precipitation and temperature. However, for the ecozone of Prairies, the correlation between NDVI and temperature was negative, which means that in this area vegetation grows worse under higher temperature condition. Generally speaking, the value of NDVI reflects the ecosystem dynamics over a long time interval under the integrated effect that depends on many parameters such as temperature and precipitation variations, soil moisture, vegetation types, and so on. That is to say, the impact of temperature and precipitation regime varies with different ecosystems, which can be verified with the above results. The arguments may partially explain relatively small values of correlation coefficients reported in Table 1. It also indicates that the deeper answers to the vegetation dynamics should be based on comparison of satellite vegetation indices with model results derived from a comprehensive land surface/ecosystem model through the assimilation of atmospheric forcing factors such as temperature, humidity, wind, precipitation, etc. Consequently, our further work is to find more comprehensive scientific understanding using land ecosystem models.

As presented above, the correspondence between NDVI and drought indices varies with ecozones. For most ecozones, the trend of annual changes in NDVI corresponded well with VPD. VPD is a major controlling factor for surface ET. Generally, high T increases both VPD and subsequent ET and leads to a drier environment, eventually reducing NPP.43 However, because the nonlinear function of increasing temperature produces a magnification effect, a same increase in air temperature will result in much higher increases in VPD at higher temperature bases than at lower temperatures. Because the average temperature in Canada is relatively low, the increase in VPD with increasing temperature is therefore relatively limited. Consequently, increased temperatures encourage vegetation growth in most regions of Canada. This finding is corroborated by other studies. For example, Chen et al.51 found that for northern high latitudes (>47.5°N), a warming climate lengthens the growing season and promotes plant growth. Zhao and Running43 also concluded that although interannual NPP was negatively correlated with air temperature over vegetated land globally from 2000 to 2009, the warming climate during this period steadily increased NPP in high-latitude, high-elevation areas. However, our study found that in the ecozones of Prairies, annual changes of NDVI correspond well with the changes in PDSI but are converse to the VPD change pattern. In the Prairies, as the average annual temperature is high (>4°C) [Fig. 2(b)] while the average annual precipitation is the lowest (<400mm) [Fig. 2(c)], the increase in VPD with increasing temperature is therefore relatively high. That is to say, increased temperature with scarce precipitation suppresses vegetation growth in the Prairies, which is different from most other regions in Canada. Consequently, the trend of annual changes in NDVI is contrary to that of VPD, while it corresponds well with the trend of PDSI which measure environmental water stress by combining information from both evaporation and precipitation.

MODIS NDVI measurements of Canadian vegetation dynamics over the past decade provided a dynamic picture of interannual variation in land surface patterns. Although some significant changes took place in both Boreal Shield and Boreal Plains for years 2002 and 2009, most forest land shows relatively little change during the period of 2000 to 2009, especially comparing with vegetation in the ecozone of Pacific Maritime, Prairies, and Taiga Shield. Vegetation in the ecozone of Pacific Maritime, Prairies, and Taiga Shield shows more dynamics than in the other ecozones. The most negative anomalies in Pacific Maritime, Prairies, and Taiga Shield are in 2007, 2001, and 2000, respectively. In general, NDVI anomalies in 2002, 2009, and 2000 over most of the area were negative, indicating region-wide drought conditions for vegetation growth, particularly in 2002. In contrast, the 2005, 2006, and 2001 anomalies over most of Canada were positive, indicating above-normal greening with favorable conditions for vegetation growth nationwide, particularly in 2005. The analyzed annual changes in NDVI averaged over different ecozones further indicated stressed vegetation conditions in 2002 and 2009 across Canada and that both 2005 and 2006 were favorable years for vegetation growth nationwide. Furthermore, vegetation dynamics over the past decade varied across regions, as shown in the spatially distributed maps of growing-season NDVI anomalies.

The impacts of climatic variability on vegetation growth during this period were also investigated. Although the correlations between NDVI and climatic variables varied with ecozones, a higher correlation was found between NDVI and temperature than between NDVI and precipitation in most ecozones. However, for the ecozone of Taiga Plains, the correlation between NDVI and temperature was much lower than that between NDVI and precipitation. In addition, almost all the correlations between NDVI and climatic variables were positive, suggesting that vegetation grows better with higher precipitation and temperature. However, the ecozone of Prairies is an exception where the correlation between NDVI and temperature was negative, suggesting that in this arid area vegetation grows worse under higher temperature condition. The comparisons between annual NDVI and drought indices indicated that the trend of annual changes in NDVI corresponded well with the change in VPD in most ecozones across Canada.

We are grateful for financial support provided by the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant and the NSERC Strategic Network (ForValueNet), the NSF-China Project (41201027), and a Postdoctoral Fellowship to X. Fang from the University of Quebec at Montreal. We thank M. Zhao for providing the drought indices data (PDSI and VPD).

Martínez  B., Gilabert  M. A., “Vegetation dynamics from NDVI time series analysis using the wavelet transform,” Remote Sens. Environ.. 113, (9 ), 1823 –1842 (2009), CrossRef. 0034-4257 
Anderson  L. O. et al., “Remote sensing detection of droughts in Amazonian forest canopies,” New Phytol.. 187, (3 ), 733 –750 (2010), CrossRef. 0028-646X 
Boyd  D. S., Foody  G. M., “An overview of recent remote sensing and GIS based research in ecological informatics,” Ecol. Inf.. 6, (1 ), 25 –36 (2011), CrossRef. 1574-9541 
Rundell  P. W. et al., “Environmental sensor networks in ecological research,” New Phytol.. 182, , 589 –607 (2009), CrossRef. 0028-646X 
Rouse  J. W. et al., “Monitoring vegetation systems in the Great Plains with ERTS,” in  3rd ERTS Symposium , pp. 309 –317,  U.S. Government Printing Office ,  Washington, DC  (1973).
de Beurs  K. M., Townsend  P. A., “Estimating the effect of gypsy moth defoliation using MODIS,” Remote Sens. Environ.. 112, (10 ), 3983 –3990 (2008), CrossRef. 0034-4257 
Jin  S., Sader  S. A., “MODIS time-series imagery for forest disturbance detection and quantification of patch size effects,” Remote Sens. Environ.. 99, (4 ), 462 –470 (2005), CrossRef. 0034-4257 
Li  M. et al., “Investigating phenological changes using MODIS vegetation indices in deciduous broadleaf forest over continental U.S. during 2000–2008,” Ecol. Inf.. 5, (5 ), 410 –417 (2010), CrossRef. 1574-9541 
Spruce  J. P. et al., “Assessment of MODIS NDVI time series data products for detecting forest defoliation by gypsy moth outbreaks,” Remote Sens. Environ.. 115, (2 ), 427 –437 (2011), CrossRef. 0034-4257 
Yuhas  A. N., Scuderi  L. A., “MODIS-derived NDVI characterisation of drought-induced evergreen dieoff in Western North America,” Geogr. Res.. 47, (1 ), 34 –45 (2009), CrossRef. 1745-5863 
Narasimhan  R., Stow  D., “Daily MODIS products for analyzing early season vegetation dynamics across the North Slope of Alaska,” Remote Sens. Environ.. 114, (6 ), 1251 –1262 (2010), CrossRef. 0034-4257 
Wang  Q. et al., “Similarities in ground- and satellite-based NDVI time series and their relationship to physiological activity of a Scots pine forest in Finland,” Remote Sens. Environ.. 93, (1–2 ), 225 –237 (2004), CrossRef. 0034-4257 
Eklundh  L. et al., “Mapping insect defoliation in Scots pine with MODIS time-series data,” Remote Sens. Environ.. 113, (7 ), 1566 –1573 (2009), CrossRef. 0034-4257 
Jönsson  A. M. et al., “Annual changes in MODIS vegetation indices of Swedish coniferous forests in relation to snow dynamics and tree phenology,” Remote Sens. Environ.. 114, (11 ), 2719 –2730 (2010), CrossRef. 0034-4257 
Verbesselt  J. et al., “Forecasting tree mortality using change metrics derived from MODIS satellite data,” For. Ecol. Manage.. 258, (7 ), 1166 –1173 (2009), CrossRef. 0378-1127 
Jepsen  J. U. et al., “Monitoring the spatio-temporal dynamics of geometrid moth outbreaks in birch forest using MODIS-NDVI data,” Remote Sens. Environ.. 113, (9 ), 1939 –1947 (2009), CrossRef. 0034-4257 
Julien  Y., Sobrino  J. A., “Comparison of cloud-reconstruction methods for time series of composite NDVI data,” Remote Sens. Environ.. 114, (3 ), 618 –625 (2010), CrossRef. 0034-4257 
Myneni  R. et al., “Increased plant growth in the northern high latitudes from 1981 to 1991,” Nature. 386, , 698 –702 (1997), CrossRef. 0028-0836 
Badeck  F.-W. et al., “Responses of spring phenology to climate change,” New Phytol.. 162, (2 ), 295 –309 (2004), CrossRef. 0028-646X 
Menzel  A. et al., “European phenological response to climate change matches the warming pattern,” Global Change Biol.. 12, (10 ), 1969 –1976 (2006), CrossRef. 1354-1013 
Schwarta  M. D. et al., “Onset of spring starting earlier across the Northern Hemisphere,” Global Change Biol.. 12, (2 ), 343 –351 (2006), CrossRef. 1354-1013 
Wang  X. et al., “Spring temperature change and its implication in the change of vegetation growth in North America from 1982 to 2006,” Proc. Natl. Acad. Sci. U. S. A.. 108, (4 ), 1240 –1245 (2011), CrossRef. 0027-8424 
Yu  H. et al., “Winter and spring warming result in delayed spring phenology on the Tibetan Plateau,” Proc. Natl. Acad. Sci. U. S. A.. 107, (51 ), 22151 –22156 (2010), CrossRef. 0027-8424 
Chen  H. et al., “Delayed spring phenology on the Tibetan Plateau may also be attributable to other factors than winter and spring warming,” Proc. Natl. Acad. Sci. U. S. A.. 108, (19 ), E93  (2011), CrossRef. 0027-8424 
Goetz  S. J. et al., “Satellite-observed photosynthetic trends across boreal North America associated with climate and fire disturbance,” Proc. Natl. Acad. Sci. U. S. A.. 102, (38 ), 13521 –13525 (2005), CrossRef. 0027-8424 
Anyamba  A., Tucker  C. J., “Analysis of Sahelian vegetation dynamics using NOAA-AVHRR NDVI data from 1981–2003,” J. Arid Environ.. 63, (3 ), 596 –614 (2005), CrossRef. 0140-1963 
Xu  L. et al., “Widespread decline in greenness of Amazonian vegetation due to the 2010 drought,” Geophys. Res. Lett.. 38, , L07402  (2011), CrossRef. 0094-8276 
Coops  N. C. et al., “Assessment of QuickBird high spatial resolution imagery to detect red attack damage due to mountain pine beetle infestation,” Remote Sens. Environ.. 103, (1 ), 67 –80 (2006), CrossRef. 0034-4257 
Wulder  M. A. et al., “Multi-temporal analysis of high spatial resolution imagery for disturbance monitoring,” Remote Sens. Environ.. 112, (6 ), 2729 –2740 (2008), CrossRef. 0034-4257 
White  J. C. et al., “Detection of red attack stage mountain pine beetle infestation with high spatial resolution satellite imagery,” Remote Sens. Environ.. 96, (3–4 ), 340 –351 (2005), CrossRef. 0034-4257 
Wulder  M. A. et al., “Monitoring the impacts of mountain pine beetle mitigation,” For. Ecol. Manage.. 258, (7 ), 1181 –1187 (2009), CrossRef. 0378-1127 
Coops  N. C. et al., “Integrating remotely sensed and ancillary data sources to characterize a mountain pine beetle infestation,” Remote Sens. Environ.. 105, (2 ), 83 –97 (2006), CrossRef. 0034-4257 
Wulder  M. A. et al., “Estimating the probability of mountain pine beetle red-attack damage,” Remote Sens. Environ.. 101, (2 ), 150 –166 (2006), CrossRef. 0034-4257 
Goodwin  N. R. et al., “Estimation of insect infestation dynamics using a temporal sequence of Landsat data,” Remote Sens. Environ.. 112, (9 ), 3680 –3689 (2008), CrossRef. 0034-4257 
Coops  N. C. et al., “Assessing changes in forest fragmentation following infestation using time series Landsat imagery,” For. Ecol. Manage.. 259, (12 ), 2355 –2365 (2010), CrossRef. 0378-1127 
Coops  N. C. et al., “Prediction and assessment of bark beetle-induced mortality of lodgepole pine using estimates of stand vigor derived from remotely sensed data,” Remote Sens. Environ.. 113, (5 ), 1058 –1066 (2009), CrossRef. 0034-4257 
Coops  N. C., Wulder  M. A., “Estimating the reduction in gross primary production due to mountain pine beetle infestation using satellite observations,” Int. J. Remote Sens.. 31, (8 ), 2129 –2138 (2010), CrossRef. 0143-1161 
Coops  N. C. et al., “Large area monitoring with a MODIS-based Disturbance Index (DI) sensitive to annual and seasonal variations,” Remote Sens. Environ.. 113, (6 ), 1250 –1261 (2009), CrossRef. 0034-4257 
Hilker  T. et al., “A new data fusion model for high spatial- and temporal-resolution mapping of forest disturbance based on Landsat and MODIS,” Remote Sens. Environ.. 113, (8 ), 1613 –1627 (2009), CrossRef. 0034-4257 
Wulder  M. A. et al., “Multiscale satellite and spatial information and analysis framework in support of a large-area forest monitoring and inventory update,” Environ. Monit. Assess.. 170, (1–4 ), 1 –17 (2009), CrossRef. 0167-6369 
Fraser  R. H., Latifovic  R., “Mapping insect-induced tree defoliation and mortality using coarse spatial resolution satellite imagery,” Int. J. Remote Sens.. 26, (1 ), 193 –200 (2005), CrossRef. 0143-1161 
Samanta  A. et al., “Amazon forests did not green-up during the 2005 drought,” Geophys. Res. Lett.. 37, , L05401  (2010), CrossRef. 0094-8276 
Zhao  M., Running  S. W., “Drought-induced reduction in global terrestrial net primary production from 2000 through 2009,” Science. 329, (5994 ), 940 –943 (2010), CrossRef. 0036-8075 
Prasad  V. K. et al., “Spatial patterns of vegetation phenology metrics and related climatic controls of eight contrasting forest types in India-analysis from remote sensing datasets,” Theor. Appl. Climatol.. 89, , 95 –107 (2007), CrossRef. 1434-4483 
Bonsal  B. R., Wheaton  E. E., “Atmospheric circulation comparisons between the 2001 and 2002 and the 1961 and 1988 Canadian prairie droughts,” Atmos.-Ocean. 43, (2 ), 163 –172 (2005), CrossRef. 0705-5900 
Girardin  M. P. et al., “Trends and periodicities in the Canadian drought code and their relationships with atmospheric circulation for the southern Canadian boreal forest,” Can. J. For. Res.. 34, (1 ), 103 –119 (2004), CrossRef. 0045-5067 
Stewart  R., Lawford  R., The 1999-2005 Canadian Prairies Drought: Science, Impacts, and Lessons. ,  The Drought Research Initiative ,  Winnipeg, Montitoba, California  (2011).
Hogg  E. H. T. et al., “Impacts of a regional drought on the productivity, dieback, and biomass of western Canadian aspen forests,” Can. J. For. Res.. 38, (6 ), 1373 –1384 (2008), CrossRef. 0045-5067 
Peng  C. H. et al., “A drought-induced pervasive increase in tree mortality across Canada’s boreal forests,” Nat. Clim. Change. 1, , 467 –471 (2011), CrossRef. 1758-678X 
Hanesiak  J. M. et al., “Characterization and summary of the 1999–2005 Canadian Prairie Drought,” Atmos.-Ocean. 49, (4 ), 421 –452 (2011), CrossRef. 0705-5900 
Chen  J. M. et al., “Boreal ecosystems sequestered more carbon in warmer years,” Geophys. Res. Lett.. 33, (10 ), L10803  (2006), CrossRef. 0094-8276 

Xiuqin Fang received her BSc and MSc degrees in cartography and geographic information science from Nanjing University, China, and her PhD degree in hydrology and water resources in Hohai Univertisy, China. From 2010 to 2012, she worked on drought study as a postdoc at the University of Quebec at Montreal, Canada. Now she is an associate professor in the School of Earth Sciences and Engineering, Hohai University. Her research interests include distributed hydrologic modeling and application of remote sensing.

Qiuan Zhu is an associate professor in the Laboratory for Ecological Forecasting and Global Change Northwest Agriculture and Forest University, China. He received his BS degree in mathematics and applied mathematics from Hohai University, China. He received his MSc and PhD degrees in geography from Nanjing University, China. His area of specialization is hydrological and ecological modeling.

Huai Chen is a professor in the Laboratory for Ecological Forecasting and Global Change, Northwest Agriculture and Forest University, China. He received his BA degree in environment engineering from Chongqing University, China. He received his MSc degree in ecology and his PhD degree in plant ecology from Chengdu Institute of Biology, Chinese Academy of Sciences, China. His research interests are global change ecology and biology and carbon biogeochemical cycles in/among natural ecosystems.

Zhihai Ma received his BS and MS degrees in forest sciences from Northeast Forestry University, China. He received the MS degree in science of applied statistics and the PhD degree in forest science from Syracuse University, New York.

Weifeng Wang received his BSc degree in environmental science from Central South Forestry University, China, MSc degree in forest management in Chinese Academy of Forestry, and PhD degree in sciences of environment at the University of Quebec at Montreal, Canada. His research interests include global change ecology, ecological modeling, and forest ecology and management.

Xinzhang Song is an associate professor at Zhejiang A&F University, China. He obtained his PhD degree in ecology from the Chinese Academy of Forestry in 2007. His postdoctoral work took him to the University of Quebec at Montreal, where he studied the carbon and nitrogen cycling of terrestrial ecosystems. His current research interests include global change ecology and carbon and nitrogen dynamics of forest ecosystem.

Pengxiang Zhao is an associate professor in the Laboratory for Ecological Forecasting and Global Change, Northwest Agriculture and Forest University, China. His research interest is forest ecosystems.

Changhui Peng is a professor and the director for Ecological Modeling and Carbon Science Lab, Institute of Environment Sciences, Department of Biology Sciences, University of Quebec at Montreal, Montreal, Canada. He received his PhD degree in faculty of science from the University of Aix-Marseille III, France. His research interests are global carbon cycle and earth system modeling.

© The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

Citation

Xiuqin Fang ; Qiuan Zhu ; Huai Chen ; Zhihai Ma ; Weifeng Wang, et al.
"Analysis of vegetation dynamics and climatic variability impacts on greenness across Canada using remotely sensed data from 2000 to 2009", J. Appl. Remote Sens. 8(1), 083666 (Mar 17, 2014). ; http://dx.doi.org/10.1117/1.JRS.8.083666


Figures

Graphic Jump LocationF1 :

Study area in Canadian territory.

Graphic Jump LocationF2 :

Average annual normalized difference vegetation index (NDVI) (a) derived from MODIS 16-day NDVI products, temperature (b), and precipitation (c) derived from National Centre for Environmental Prediction (NCEP).

Graphic Jump LocationF3 :

NDVI anomalies in growing seasons from 2000 to 2009.

Graphic Jump LocationF4 :

Annual changes in NDVI, PDSI, and VPD averaged within each of the 13 ecozones.

Tables

Table Grahic Jump Location
Table 1Correlations between average annual NDVI and climatic variables with different target areas; + slope indicates a positive relationship and − slope indicates a negative relationship.
Table Grahic Jump Location
Table 2Area percentages of drought conditions and favorable conditions (%).

References

Martínez  B., Gilabert  M. A., “Vegetation dynamics from NDVI time series analysis using the wavelet transform,” Remote Sens. Environ.. 113, (9 ), 1823 –1842 (2009), CrossRef. 0034-4257 
Anderson  L. O. et al., “Remote sensing detection of droughts in Amazonian forest canopies,” New Phytol.. 187, (3 ), 733 –750 (2010), CrossRef. 0028-646X 
Boyd  D. S., Foody  G. M., “An overview of recent remote sensing and GIS based research in ecological informatics,” Ecol. Inf.. 6, (1 ), 25 –36 (2011), CrossRef. 1574-9541 
Rundell  P. W. et al., “Environmental sensor networks in ecological research,” New Phytol.. 182, , 589 –607 (2009), CrossRef. 0028-646X 
Rouse  J. W. et al., “Monitoring vegetation systems in the Great Plains with ERTS,” in  3rd ERTS Symposium , pp. 309 –317,  U.S. Government Printing Office ,  Washington, DC  (1973).
de Beurs  K. M., Townsend  P. A., “Estimating the effect of gypsy moth defoliation using MODIS,” Remote Sens. Environ.. 112, (10 ), 3983 –3990 (2008), CrossRef. 0034-4257 
Jin  S., Sader  S. A., “MODIS time-series imagery for forest disturbance detection and quantification of patch size effects,” Remote Sens. Environ.. 99, (4 ), 462 –470 (2005), CrossRef. 0034-4257 
Li  M. et al., “Investigating phenological changes using MODIS vegetation indices in deciduous broadleaf forest over continental U.S. during 2000–2008,” Ecol. Inf.. 5, (5 ), 410 –417 (2010), CrossRef. 1574-9541 
Spruce  J. P. et al., “Assessment of MODIS NDVI time series data products for detecting forest defoliation by gypsy moth outbreaks,” Remote Sens. Environ.. 115, (2 ), 427 –437 (2011), CrossRef. 0034-4257 
Yuhas  A. N., Scuderi  L. A., “MODIS-derived NDVI characterisation of drought-induced evergreen dieoff in Western North America,” Geogr. Res.. 47, (1 ), 34 –45 (2009), CrossRef. 1745-5863 
Narasimhan  R., Stow  D., “Daily MODIS products for analyzing early season vegetation dynamics across the North Slope of Alaska,” Remote Sens. Environ.. 114, (6 ), 1251 –1262 (2010), CrossRef. 0034-4257 
Wang  Q. et al., “Similarities in ground- and satellite-based NDVI time series and their relationship to physiological activity of a Scots pine forest in Finland,” Remote Sens. Environ.. 93, (1–2 ), 225 –237 (2004), CrossRef. 0034-4257 
Eklundh  L. et al., “Mapping insect defoliation in Scots pine with MODIS time-series data,” Remote Sens. Environ.. 113, (7 ), 1566 –1573 (2009), CrossRef. 0034-4257 
Jönsson  A. M. et al., “Annual changes in MODIS vegetation indices of Swedish coniferous forests in relation to snow dynamics and tree phenology,” Remote Sens. Environ.. 114, (11 ), 2719 –2730 (2010), CrossRef. 0034-4257 
Verbesselt  J. et al., “Forecasting tree mortality using change metrics derived from MODIS satellite data,” For. Ecol. Manage.. 258, (7 ), 1166 –1173 (2009), CrossRef. 0378-1127 
Jepsen  J. U. et al., “Monitoring the spatio-temporal dynamics of geometrid moth outbreaks in birch forest using MODIS-NDVI data,” Remote Sens. Environ.. 113, (9 ), 1939 –1947 (2009), CrossRef. 0034-4257 
Julien  Y., Sobrino  J. A., “Comparison of cloud-reconstruction methods for time series of composite NDVI data,” Remote Sens. Environ.. 114, (3 ), 618 –625 (2010), CrossRef. 0034-4257 
Myneni  R. et al., “Increased plant growth in the northern high latitudes from 1981 to 1991,” Nature. 386, , 698 –702 (1997), CrossRef. 0028-0836 
Badeck  F.-W. et al., “Responses of spring phenology to climate change,” New Phytol.. 162, (2 ), 295 –309 (2004), CrossRef. 0028-646X 
Menzel  A. et al., “European phenological response to climate change matches the warming pattern,” Global Change Biol.. 12, (10 ), 1969 –1976 (2006), CrossRef. 1354-1013 
Schwarta  M. D. et al., “Onset of spring starting earlier across the Northern Hemisphere,” Global Change Biol.. 12, (2 ), 343 –351 (2006), CrossRef. 1354-1013 
Wang  X. et al., “Spring temperature change and its implication in the change of vegetation growth in North America from 1982 to 2006,” Proc. Natl. Acad. Sci. U. S. A.. 108, (4 ), 1240 –1245 (2011), CrossRef. 0027-8424 
Yu  H. et al., “Winter and spring warming result in delayed spring phenology on the Tibetan Plateau,” Proc. Natl. Acad. Sci. U. S. A.. 107, (51 ), 22151 –22156 (2010), CrossRef. 0027-8424 
Chen  H. et al., “Delayed spring phenology on the Tibetan Plateau may also be attributable to other factors than winter and spring warming,” Proc. Natl. Acad. Sci. U. S. A.. 108, (19 ), E93  (2011), CrossRef. 0027-8424 
Goetz  S. J. et al., “Satellite-observed photosynthetic trends across boreal North America associated with climate and fire disturbance,” Proc. Natl. Acad. Sci. U. S. A.. 102, (38 ), 13521 –13525 (2005), CrossRef. 0027-8424 
Anyamba  A., Tucker  C. J., “Analysis of Sahelian vegetation dynamics using NOAA-AVHRR NDVI data from 1981–2003,” J. Arid Environ.. 63, (3 ), 596 –614 (2005), CrossRef. 0140-1963 
Xu  L. et al., “Widespread decline in greenness of Amazonian vegetation due to the 2010 drought,” Geophys. Res. Lett.. 38, , L07402  (2011), CrossRef. 0094-8276 
Coops  N. C. et al., “Assessment of QuickBird high spatial resolution imagery to detect red attack damage due to mountain pine beetle infestation,” Remote Sens. Environ.. 103, (1 ), 67 –80 (2006), CrossRef. 0034-4257 
Wulder  M. A. et al., “Multi-temporal analysis of high spatial resolution imagery for disturbance monitoring,” Remote Sens. Environ.. 112, (6 ), 2729 –2740 (2008), CrossRef. 0034-4257 
White  J. C. et al., “Detection of red attack stage mountain pine beetle infestation with high spatial resolution satellite imagery,” Remote Sens. Environ.. 96, (3–4 ), 340 –351 (2005), CrossRef. 0034-4257 
Wulder  M. A. et al., “Monitoring the impacts of mountain pine beetle mitigation,” For. Ecol. Manage.. 258, (7 ), 1181 –1187 (2009), CrossRef. 0378-1127 
Coops  N. C. et al., “Integrating remotely sensed and ancillary data sources to characterize a mountain pine beetle infestation,” Remote Sens. Environ.. 105, (2 ), 83 –97 (2006), CrossRef. 0034-4257 
Wulder  M. A. et al., “Estimating the probability of mountain pine beetle red-attack damage,” Remote Sens. Environ.. 101, (2 ), 150 –166 (2006), CrossRef. 0034-4257 
Goodwin  N. R. et al., “Estimation of insect infestation dynamics using a temporal sequence of Landsat data,” Remote Sens. Environ.. 112, (9 ), 3680 –3689 (2008), CrossRef. 0034-4257 
Coops  N. C. et al., “Assessing changes in forest fragmentation following infestation using time series Landsat imagery,” For. Ecol. Manage.. 259, (12 ), 2355 –2365 (2010), CrossRef. 0378-1127 
Coops  N. C. et al., “Prediction and assessment of bark beetle-induced mortality of lodgepole pine using estimates of stand vigor derived from remotely sensed data,” Remote Sens. Environ.. 113, (5 ), 1058 –1066 (2009), CrossRef. 0034-4257 
Coops  N. C., Wulder  M. A., “Estimating the reduction in gross primary production due to mountain pine beetle infestation using satellite observations,” Int. J. Remote Sens.. 31, (8 ), 2129 –2138 (2010), CrossRef. 0143-1161 
Coops  N. C. et al., “Large area monitoring with a MODIS-based Disturbance Index (DI) sensitive to annual and seasonal variations,” Remote Sens. Environ.. 113, (6 ), 1250 –1261 (2009), CrossRef. 0034-4257 
Hilker  T. et al., “A new data fusion model for high spatial- and temporal-resolution mapping of forest disturbance based on Landsat and MODIS,” Remote Sens. Environ.. 113, (8 ), 1613 –1627 (2009), CrossRef. 0034-4257 
Wulder  M. A. et al., “Multiscale satellite and spatial information and analysis framework in support of a large-area forest monitoring and inventory update,” Environ. Monit. Assess.. 170, (1–4 ), 1 –17 (2009), CrossRef. 0167-6369 
Fraser  R. H., Latifovic  R., “Mapping insect-induced tree defoliation and mortality using coarse spatial resolution satellite imagery,” Int. J. Remote Sens.. 26, (1 ), 193 –200 (2005), CrossRef. 0143-1161 
Samanta  A. et al., “Amazon forests did not green-up during the 2005 drought,” Geophys. Res. Lett.. 37, , L05401  (2010), CrossRef. 0094-8276 
Zhao  M., Running  S. W., “Drought-induced reduction in global terrestrial net primary production from 2000 through 2009,” Science. 329, (5994 ), 940 –943 (2010), CrossRef. 0036-8075 
Prasad  V. K. et al., “Spatial patterns of vegetation phenology metrics and related climatic controls of eight contrasting forest types in India-analysis from remote sensing datasets,” Theor. Appl. Climatol.. 89, , 95 –107 (2007), CrossRef. 1434-4483 
Bonsal  B. R., Wheaton  E. E., “Atmospheric circulation comparisons between the 2001 and 2002 and the 1961 and 1988 Canadian prairie droughts,” Atmos.-Ocean. 43, (2 ), 163 –172 (2005), CrossRef. 0705-5900 
Girardin  M. P. et al., “Trends and periodicities in the Canadian drought code and their relationships with atmospheric circulation for the southern Canadian boreal forest,” Can. J. For. Res.. 34, (1 ), 103 –119 (2004), CrossRef. 0045-5067 
Stewart  R., Lawford  R., The 1999-2005 Canadian Prairies Drought: Science, Impacts, and Lessons. ,  The Drought Research Initiative ,  Winnipeg, Montitoba, California  (2011).
Hogg  E. H. T. et al., “Impacts of a regional drought on the productivity, dieback, and biomass of western Canadian aspen forests,” Can. J. For. Res.. 38, (6 ), 1373 –1384 (2008), CrossRef. 0045-5067 
Peng  C. H. et al., “A drought-induced pervasive increase in tree mortality across Canada’s boreal forests,” Nat. Clim. Change. 1, , 467 –471 (2011), CrossRef. 1758-678X 
Hanesiak  J. M. et al., “Characterization and summary of the 1999–2005 Canadian Prairie Drought,” Atmos.-Ocean. 49, (4 ), 421 –452 (2011), CrossRef. 0705-5900 
Chen  J. M. et al., “Boreal ecosystems sequestered more carbon in warmer years,” Geophys. Res. Lett.. 33, (10 ), L10803  (2006), CrossRef. 0094-8276 

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Related Book Chapters

Topic Collections

PubMed Articles
Advertisement


 

  • Don't have an account?
  • Subscribe to the SPIE Digital Library
  • Create a FREE account to sign up for Digital Library content alerts and gain access to institutional subscriptions remotely.
Access This Article
Sign in or Create a personal account to Buy this article ($20 for members, $25 for non-members).
Access This Proceeding
Sign in or Create a personal account to Buy this article ($15 for members, $18 for non-members).
Access This Chapter

Access to SPIE eBooks is limited to subscribing institutions and is not available as part of a personal subscription. Print or electronic versions of individual SPIE books may be purchased via SPIE.org.