Characterization of land surface phenology of land cover types on the Flooding
            Pampa (Argentina): an approach using NDVI time series

Due to a rapid increase in accessible Earth observation data coupled with high computing and storage capabilities, multiple efforts over the past few years have aimed to map land use/land cover using image time series with promising outcomes. Here, we evaluate the comparative performance of alternative land cover classifications generated by using only (1) phenological metrics derived from either of two land surface phenology models, or (2) a suite of spectral band percentiles and normalized ratios (spectral variables), or (3) a combination of phenological metrics and spectral variables. First, several annual time series of remotely sensed data were assembled: Accumulated growing degree-days (AGDD) from the MODerate resolution Imaging Spectroradiometer (MODIS) 8-day land surface temperature products, 2-band Enhanced Vegetation Index (EVI2), and the spectral variables from the Harmonized Landsat Sentinel-2, as well as from the U.S. Landsat Analysis Ready Data surface reflectance products. Then, at each pixel, EVI2 time series were fitted using two different land surface phenology models: The Convex Quadratic model (CxQ), in which EVI2 = f(AGDD) and the Hybrid Piecewise Logistic Model (HPLM), in which EVI2 = f(day of year). Phenometrics and spectral variables were submitted separately and together to Random Forest Classifiers (RFC) to depict land use/land cover in Roberts County, South Dakota. HPLM RFC models showed slightly better accuracy than CxQ RFC models (about 1% relative higher in overall accuracy). Compared to phenometrically-based RFC models, spectrally-based RFC models yielded more accurate land cover maps, especially for non-crop cover types. However, the RFC models built from spectral variables could not accurately classify the wheat class, which contained mostly spring wheat with some fields in durum or winter varieties. The most accurate RFC models were obtained when using both phenometrics and spectral variables as inputs. The combined-variable RFC models overcame weaknesses of both phenometrically-based classification (low accuracy for non-vegetated covers) and spectrally-based classification (low accuracy for wheat). The analysis of important variables indicated that land cover classification for this study area was strongly driven by variables related to the initial green-up phase of seasonal growth and maximum fitted EVI2. For a deeper evaluation of RFC performance, RFC classifications were also executed with several alternative sampling scenarios, including different spatiotemporal filters to improve accuracy of sample pools and different sample sizes. Results indicated that a sample pool with less filtering yielded the most accurate predicted land cover map and a stratified random sample dataset covering approximately 0.25% or more of the study area were required to achieve an accurate land cover map. In case of data scarcity, a smaller dataset might be acceptable, but should not smaller than 0.05% of the study area.

Download Full-text

Spatiotemporal Variability of Land Surface Albedo over the Tibet Plateau from 2001 to 2019

Remote Sensing ◽

10.3390/rs12071188 ◽

2020 ◽

Vol 12 (7) ◽

pp. 1188

Author(s):

Xingwen Lin ◽

Jianguang Wen ◽

Qinhuo Liu ◽

Dongqin You ◽

Shengbiao Wu ◽

...

Keyword(s):

Climate Change ◽

Time Series ◽

Land Cover ◽

Snow Cover ◽

Land Surface ◽

Radiative Forcing ◽

Surface Albedo ◽

Tibet Plateau ◽

Land Cover Types ◽

Yearly Average

As an essential climate variable (ECV), land surface albedo plays an important role in the Earth surface radiation budget and regional or global climate change. The Tibetan Plateau (TP) is a sensitive environment to climate change, and understanding its albedo seasonal and inter-annual variations is thus important to help capture the climate change rules. In this paper, we analyzed the large-scale spatial patterns, temporal trends, and seasonal variability of land surface albedo overall the TP, based on the moderate resolution imaging spectroradiometer (MODIS) MCD43 albedo products from 2001 to 2019. Specifically, we assessed the correlations between the albedo anomaly and the anomalies of normalized difference vegetation index (NDVI), the fraction of snow cover (snow cover), and land surface temperature (LST). The results show that there are larger albedo variations distributed in the mountainous terrain of the TP. Approximately 10.06% of the land surface is identified to have been influenced by the significant albedo variation from the year 2001 to 2019. The yearly averaged albedo was decreased significantly at a rate of 0.0007 (Sen’s slope) over the TP. Additionally, the yearly average snow cover was decreased at a rate of 0.0756. However, the yearly average NDVI and LST were increased with slopes of 0.0004 and 0.0253 over the TP, respectively. The relative radiative forcing (RRF) caused by the land cover change (LCC) is larger than that caused by gradual albedo variation in steady land cover types. Overall, the RRF due to gradual albedo variation varied from 0.0005 to 0.0170 W/m2, and the RRF due to LCC variation varied from 0.0037 to 0.0243 W/m2 during the years 2001 to 2019. The positive RRF caused by gradual albedo variation or the LCC can strengthen the warming effects in the TP. The impact of the gradual albedo variations occurring in the steady land cover types was very low between 2001 and 2019 because the time series was short, and it therefore cannot be neglected when examining radiative forcing for a long time series regarding climate change.

Download Full-text

Influence of Varying Solar Zenith Angles on Land Surface Phenology Derived from Vegetation Indices: A Case Study in the Harvard Forest

Remote Sensing ◽

10.3390/rs13204126 ◽

2021 ◽

Vol 13 (20) ◽

pp. 4126

Author(s):

Yang Li ◽

Ziti Jiao ◽

Kaiguang Zhao ◽

Yadong Dong ◽

Yuyu Zhou ◽

...

Keyword(s):

Time Series ◽

Land Surface ◽

Vegetation Index ◽

Vegetation Indices ◽

Mean Difference ◽

Land Surface Phenology ◽

Harvard Forest ◽

Ndvi Time Series ◽

Bidirectional Reflectance Distribution

Vegetation indices are widely used to derive land surface phenology (LSP). However, due to inconsistent illumination geometries, reflectance varies with solar zenith angles (SZA), which in turn affects the vegetation indices, and thus the derived LSP. To examine the SZA effect on LSP, the MODIS bidirectional reflectance distribution function (BRDF) product and a BRDF model were employed to derive LSPs under several constant SZAs (i.e., 0°, 15°, 30°, 45°, and 60°) in the Harvard Forest, Massachusetts, USA. The LSPs derived under varying SZAs from the MODIS nadir BRDF-adjusted reflectance (NBAR) and MODIS vegetation index products were used as baselines. The results show that with increasing SZA, NDVI increases but EVI decreases. The magnitude of SZA-induced NDVI/EVI changes suggests that EVI is more sensitive to varying SZAs than NDVI. NDVI and EVI are comparable in deriving the start of season (SOS), but EVI is more accurate when deriving the end of season (EOS). Specifically, NDVI/EVI-derived SOSs are relatively close to those derived from ground measurements, with an absolute mean difference of 8.01 days for NDVI-derived SOSs and 9.07 days for EVI-derived SOSs over ten years. However, a considerable lag exists for EOSs derived from vegetation indices, especially from the NDVI time series, with an absolute mean difference of 14.67 days relative to that derived from ground measurements. The SOSs derived from NDVI time series are generally earlier, while those from EVI time series are delayed. In contrast, the EOSs derived from NDVI time series are delayed; those derived from the simulated EVI time series under a fixed illumination geometry are also delayed, but those derived from the products with varying illumination geometries (i.e., MODIS NBAR product and MODIS vegetation index product) are advanced. LSPs derived from varying illumination geometries could lead to a difference spanning from a few days to a month in this case study, which highlights the importance of normalizing the illumination geometry when deriving LSP from NDVI/EVI time series.

Download Full-text

Discriminating the Mediterranean Pinus spp. using the land surface phenology extracted from the whole MODIS NDVI time series and machine learning algorithms

Earth Resources and Environmental Remote Sensing/GIS Applications VIII ◽

10.1117/12.2278557 ◽

2017 ◽

Author(s):

Victor F. Rodriguez-Galiano ◽

David Aragones ◽

Jose A. Caparros-Santiago ◽

Rafael M. Navarro-Cerrillo

Keyword(s):

Machine Learning ◽

Time Series ◽

Land Surface ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Land Surface Phenology ◽

Modis Ndvi ◽

Ndvi Time Series ◽

The Mediterranean ◽

Pinus Spp

Download Full-text

Are Vegetation Dynamics Impacted from a Nuclear Disaster? The Case of Chernobyl Using Remotely Sensed NDVI and Land Cover Data

Land ◽

10.3390/land9110433 ◽

2020 ◽

Vol 9 (11) ◽

pp. 433

Author(s):

Alexandra Gemitzi

Keyword(s):

Time Series ◽

Land Cover ◽

Human Activities ◽

Vegetation Dynamics ◽

Positive Impact ◽

Remotely Sensed ◽

Exclusion Zone ◽

Land Cover Types ◽

Nuclear Disaster ◽

Ndvi Time Series

There is a growing interest for scientists and society to acquire deep knowledge on the impacts from environmental disasters. The present work deals with the investigation of vegetation dynamics in the Chernobyl area, a place widely known for the devastating nuclear disaster on the 26th of April 1986. To unveil any possible long-term radiation effects on vegetation phenology, the remotely sensed normalized difference vegetation index (NDVI) from the Moderate Resolution Imaging Spectroradiometer (MODIS) was analyzed within the 30 km Exclusion Zone, where all human activities were ceased at that time and public access and inhabitation have been prohibited ever since. The analysis comprised applications of seasonal trend analysis using two techniques, i.e., pixel-wise NDVI time series and spatially averaged NDVI time series. Both techniques were applied in each one of the individual land cover types. To assess the existence of abnormal vegetation dynamics, the same analyses were conducted in two broader zones, i.e., from 30 to 60 km and from 60 to 90 km, away from Chernobyl area, where human activities were not substantially altered. Results of both analyses indicated that vegetation dynamics in the 30 km Exclusion Zone correspond to increasing plant productivity at a rate considerably higher than that of the other two examined zones. The outcome of the analyses presented herein attributes greening trends in the 30 km and the 30 to 60 km zones to a combination of climate, minimized human impact and a consequent prevalence of land cover types which seem to be well adapted to increased radioactivity. The vegetation greening trends observed in the third zone, i.e., the 90 km zone, are indicative of the combination of climate and increasing human activities. Results indicate the positive impact from the absence of human activities on vegetation dynamics as far as vegetation productivity and phenology are concerned in the 30 km Exclusion Zone, and to a lower extent in the 60 km zone. Furthermore, there is evidence that land cover changes evolve into the prevalence of woody vegetation in an area with increased levels of radioactivity.

Download Full-text

Characterizing land cover/land use from multiple years of Landsat and MODIS time series: A novel approach using land surface phenology modeling and random forest classifier

Remote Sensing of Environment ◽

10.1016/j.rse.2018.12.016 ◽

2020 ◽

Vol 238 ◽

pp. 111017 ◽

Cited By ~ 17

Author(s):

Lan H. Nguyen ◽

Deepak R. Joshi ◽

David E. Clay ◽

Geoffrey M. Henebry

Keyword(s):

Land Use ◽

Time Series ◽

Random Forest ◽

Land Cover ◽

Land Surface ◽

Random Forest Classifier ◽

Land Surface Phenology ◽

Novel Approach

Download Full-text

Using Hidden Markov Models for Land Surface Phenology: An Evaluation Across a Range of Land Cover Types in Southeast Spain

Remote Sensing ◽

10.3390/rs11050507 ◽

2019 ◽

Vol 11 (5) ◽

pp. 507

Author(s):

Miguel García ◽

Hassane Moutahir ◽

Grant Casady ◽

Susana Bautista ◽

Francisco Rodríguez

Keyword(s):

Land Cover ◽

Spatial Variability ◽

Hidden Markov Models ◽

Land Surface ◽

Markov Models ◽

Hidden Markov ◽

Land Surface Phenology ◽

Land Cover Types ◽

Smoothing Methods ◽

Southeast Spain

Land Surface Phenology (LSP) metrics are increasingly being used as indicators of climate change impacts in ecosystems. For this purpose, it is necessary to use methods that can be applied to large areas with different types of vegetation, including vulnerable semiarid ecosystems that exhibit high spatial variability and low signal-to-noise ratio in seasonality. In this work, we evaluated the use of hidden Markov models (HMM) to extract phenological parameters from Moderate Resolution Imaging Spectroradiometer (MODIS) derived Normalized Difference Vegetation Index (NDVI). We analyzed NDVI time-series data for the period 2000–2018 across a range of land cover types in Southeast Spain, including rice croplands, shrublands, mixed pine forests, and semiarid steppes. Start of Season (SOS) and End of Season (EOS) metrics derived from HMM were compared with those obtained using well-established smoothing methods. When a clear and consistent seasonal variation was present, as was the case in the rice croplands, and when adjusting average curves, the smoothing methods performed as well as expected, with HMM providing consistent results. When spatial variability was high and seasonality was less clearly defined, as in the semiarid shrublands and steppe, the performance of the smoothing methods degraded. In these cases, the results from HMM were also less consistent, yet they were able to provide pixel-wise estimations of the metrics even when comparison methods did not.

Download Full-text

Analysis of vegetation and land cover dynamics in north-western Morocco during the last decade using MODIS NDVI time series data

Biogeosciences ◽

10.5194/bg-8-3359-2011 ◽

2011 ◽

Vol 8 (11) ◽

pp. 3359-3373 ◽

Cited By ~ 17

Author(s):

C. Höpfner ◽

D. Scherer

Keyword(s):

Time Series ◽

Land Cover ◽

Vegetation Dynamics ◽

Time Series Data ◽

Series Data ◽

Land Cover Types ◽

Modis Ndvi ◽

Ndvi Time Series ◽

Western Morocco ◽

North Western

Abstract. Vegetation phenology as well as the current variability and dynamics of vegetation and land cover, including its climatic and human drivers, are examined in a region in north-western Morocco that is nearly 22 700 km2 big. A gapless time series of Normalized Differenced Vegetation Index (NDVI) composite raster data from 29 September 2000 to 29 September 2009 is utilised. The data have a spatial resolution of 250 m and were acquired by the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor. The presented approach allows to compose and to analyse yearly land cover maps in a widely unknown region with scarce validated ground truth data by deriving phenological parameters. Results show that the high temporal resolution of 16 d is sufficient for (a) determining local land cover better than global land cover classifications of Plant Functional Types (PFT) and Global Land Cover 2000 (GLC2000) and (b) for drawing conclusions on vegetation dynamics and its drivers. Areas of stably classified land cover types (i.e. areas that did not change their land cover type) show climatically driven inter- and intra-annual variability with indicated influence of droughts. The presented approach to determine human-driven influence on vegetation dynamics caused by agriculture results in a more than ten times larger area compared with stably classified areas. Change detection based on yearly land cover maps shows a gain of high-productive vegetation (cropland) of about 259.3 km2. Statistically significant inter-annual trends in vegetation dynamics during the last decade could however not be discovered. A sequence of correlations was respectively carried out to extract the most important periods of rainfall responsible for the production of green biomass and for the extent of land cover types. Results show that mean daily precipitation from 1 October to 15 December has high correlation results (max. r2=0.85) on an intra-annual time scale to NDVI percentiles (50 %) of land cover types. Correlation results of mean daily precipitation from 16 September to 15 January and percentage of yearly classified area of each land cover type are medium up to high (max. r2=0.64). In all, an offset of nearly 1.5 months is detected between precipitation rates and NDVI values. High-productive vegetation (cropland) is proved to be mainly rain-fed. We conclude that identification, understanding and knowledge about vegetation phenology, and current variability of vegetation and land cover, as well as prediction methods of land cover change, can be improved using multi-year MODIS NDVI time series data. This study enhances the comprehension of current land surface dynamics and variability of vegetation and land cover in north-western Morocco. It especially offers a quick access when estimating the extent of agricultural lands.

Download Full-text

Analysis of vegetation and land cover dynamics in north-western Morocco during the last decade using MODIS NDVI time series data

Biogeosciences Discussions ◽

10.5194/bgd-8-3953-2011 ◽

2011 ◽

Vol 8 (2) ◽

pp. 3953-3998 ◽

Cited By ~ 2

Author(s):

C. Höpfner ◽

D. Scherer

Keyword(s):

Time Series ◽

Land Cover ◽

Vegetation Dynamics ◽

Time Series Data ◽

Series Data ◽

Land Cover Types ◽

Modis Ndvi ◽

Ndvi Time Series ◽

Western Morocco ◽

North Western

Abstract. Vegetation phenology as well as current variability and dynamics of vegetation and land cover including its climatic and human drivers are examined in a region in north-western Morocco of nearly 22 700 km2. A gapless time series of Normalized Differenced Vegetation Index (NDVI) composite raster data from 29 September 2000 to 29 September 2009 with a spatial resolution of 250 m and acquired by the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor is utilised. The presented approach allows to compose and analyse yearly land cover maps in a widely unknown region with scarce validated ground truth data by deriving phenological parameters. Results show that high temporal resolution of 16 d is sufficient (a) for determining land cover better than global land cover classifications of Plant Functional Types (PFT) and Global Land Cover 2000 (GLC2000), and (b) for drawing conclusions on vegetation dynamics and its drivers. Areas of stably classified land cover types show climatically driven inter- and intra-annual variability with indicated influence of droughts. The presented approach to determine human-driven influence on vegetation dynamics caused by agriculture results in a more than ten times larger area compared to the stably classified areas. Change detection based on yearly land cover maps shows a gain of high-productive vegetation (cropland) of about 259.3 km2. However, statistically significant inter-annual trends in vegetation dynamics during the last decade could not be discovered. A sequence of correlations was done to extract the most important period of rainfall for production of green biomass and for the extent of land cover types, respectively. Results show that mean daily precipitation from 1 October to 15 December has high correlation results (max. r2=0.85) at intra-annual time scale to NDVI percentiles (50%) of land cover types. Correlation results of mean daily precipitation from 16 September to 15 January and percentage of yearly classified area of each land cover type are medium up to high (max. r2=0.64). In all, an offset of nearly 1.5 months is detected between precipitation rates and NDVI in 16 d steps. High-productive vegetation (cropland) is proved to be mainly rain-fed. We conclude that identification, understanding and knowledge about vegetation phenology, and current variability of vegetation and land cover as well as prediction methods of land cover change can be improved using multi-year MODIS NDVI time series data. This study enhances the comprehension of current land surface dynamics and variability of vegetation and land cover in north-western Morocco offering a fast access especially for estimating the extent of agricultural lands.

Download Full-text