Exploring the Impact of Various Spectral Indices on Land Cover Change Detection Using Change Vector Analysis: A Case Study of Crete Island, Greece

Polykretis, Christos; Grillakis, Manolis; Alexakis, Dimitrios D.

doi:10.3390/rs12020319

Cited by 60 publications

(30 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The NDVI is often used for the mapping of changes in land cover [50,51] and, according to Singh et al [52], this index, used in the LULC classification, produces a significant improvement in classification accuracy. The BSI is mainly used to highlight the difference between agricultural and non-agricultural land thanks to its enhanced ability in identifying bare soil and fallow lands [53]. These additional indices are commonly used to improve the LULC classification.…”

Section: Dataset Compositionmentioning

confidence: 99%

Object-Oriented LULC Classification in Google Earth Engine Combining SNIC, GLCM, and Machine Learning Algorithms

2020

View full text Add to dashboard Cite

Google Earth Engine (GEE) is a versatile cloud platform in which pixel-based (PB) and object-oriented (OO) Land Use–Land Cover (LULC) classification approaches can be implemented, thanks to the availability of the many state-of-art functions comprising various Machine Learning (ML) algorithms. OO approaches, including both object segmentation and object textural analysis, are still not common in the GEE environment, probably due to the difficulties existing in concatenating the proper functions, and in tuning the various parameters to overcome the GEE computational limits. In this context, this work is aimed at developing and testing an OO classification approach combining the Simple Non-Iterative Clustering (SNIC) algorithm to identify spatial clusters, the Gray-Level Co-occurrence Matrix (GLCM) to calculate cluster textural indices, and two ML algorithms (Random Forest (RF) or Support Vector Machine (SVM)) to perform the final classification. A Principal Components Analysis (PCA) is applied to the main seven GLCM indices to synthesize in one band the textural information used for the OO classification. The proposed approach is implemented in a user-friendly, freely available GEE code useful to perform the OO classification, tuning various parameters (e.g., choose the input bands, select the classification algorithm, test various segmentation scales) and compare it with a PB approach. The accuracy of OO and PB classifications can be assessed both visually and through two confusion matrices that can be used to calculate the relevant statistics (producer’s, user’s, overall accuracy (OA)). The proposed methodology was broadly tested in a 154 km2 study area, located in the Lake Trasimeno area (central Italy), using Landsat 8 (L8), Sentinel 2 (S2), and PlanetScope (PS) data. The area was selected considering its complex LULC mosaic mainly composed of artificial surfaces, annual and permanent crops, small lakes, and wooded areas. In the study area, the various tests produced interesting results on the different datasets (OA: PB RF (L8 = 72.7%, S2 = 82%, PS = 74.2), PB SVM (L8 = 79.1%, S2 = 80.2%, PS = 74.8%), OO RF (L8 = 64%, S2 = 89.3%, PS = 77.9), OO SVM (L8 = 70.4, S2 = 86.9%, PS = 73.9)). The broad code application demonstrated very good reliability of the whole process, even though the OO classification process resulted, sometimes, too demanding on higher resolution data, considering the available computational GEE resources.

show abstract

Section: Dataset Compositionmentioning

confidence: 99%

Object-Oriented LULC Classification in Google Earth Engine Combining SNIC, GLCM, and Machine Learning Algorithms

2020

View full text Add to dashboard Cite

show abstract

“…On the other hand, according to the relative results by an administrative perspective, Chania prefecture was found to be most prone in erosion (Figure 9). The increasing amount of rainfall water from east to west as a result of rainfall's spatial variability [58] could be considered the main natural driving force.…”

Section: Discussionmentioning

confidence: 99%

“…The cloud mask was removed and the empty pixels were fulfilled by the values of cloud-free images of directly previous or next months. Afterwards, the Normalized Difference Vegetation Index (NDVI) was calculated in the ArcGIS environment by the following Equation [58]:…”

Section: Implementation Of Ruslementioning

confidence: 99%

Assessment of Intra-Annual and Inter-Annual Variabilities of Soil Erosion in Crete Island (Greece) by Incorporating the Dynamic “Nature” of R and C-Factors in RUSLE Modeling

et al. 2020

Self Cite

View full text Add to dashboard Cite

Under the continuously changing conditions of the environment, the exploration of spatial variability of soil erosion at a sub-annual temporal resolution, as well as the identification of high-soil loss time periods and areas, are crucial for implementing mitigation and land management interventions. The main objective of this study was to estimate the monthly and seasonal soil loss rates by water-induced soil erosion in Greek island of Crete for two recent hydrologically contrasting years, 2016 (dry) and 2019 (wet), as a result of Revised Universal Soil Loss Equation (RUSLE) modeling. The impact of temporal variability of the two dynamic RUSLE factors, namely rainfall erosivity (R) and cover management (C), was explored by using rainfall and remotely sensed vegetation data time-series of high temporal resolution. Soil, topographical, and land use/cover data were exploited to represent the other three static RUSLE factors, namely soil erodibility (K), slope length and steepness (LS) and support practice (P). The estimated rates were mapped presenting the spatio-temporal distribution of soil loss for the study area on a both intra-annual and inter-annual basis. The identification of high-loss months/seasons and areas in the island was achieved by these maps. Autumn (about 35 t ha−1) with October (about 61 t ha−1) in 2016, and winter (about 96 t ha−1) with February (146 t ha−1) in 2019 presented the highest mean soil loss rates on a seasonal and monthly, respectively, basis. Summer (0.22–0.25 t ha−1), with its including months, showed the lowest rates in both examined years. The intense monthly fluctuations of R-factor were found to be more influential on water-induced soil erosion than the more stabilized tendency of C-factor. In both years, olive groves in terms of agricultural land use and Chania prefecture in terms of administrative division, were detected as the most prone spatial units to erosion.

show abstract

“…Recent findings have indicated that the region's precipitation regime facilitates rainfall induced soil erosivity to a larger extend than was previously thought [26,27]. From a hydrogeological perspective, the island is mainly composed of pre-alpine and alpine carbonate formations and neogene and quaternary (alluvial) sediments which expedite water penetration; hence, the region exhibits limited surface water resources [28]. Crete is a well-studied region in the context of climate change impacts, mainly for its location and size, its environmental heterogeneity, but also due to its hydrological isolation from continental Greece, which makes it ideal for studying climate change impacts [21,[29][30][31].…”

Section: Case Study Regionmentioning

confidence: 96%

A Quantile Mapping Method to Fill in Discontinued Daily Precipitation Time Series

Grillakis

Polykretis

Manoudakis

et al. 2020

Water

Self Cite

View full text Add to dashboard Cite

We present and assess a method to estimate missing values in daily precipitation time series for the Mediterranean island of Crete. The method involves a quantile mapping methodology originally developed for the bias correction of climate models’ output. The overall methodology is based on a two-step procedure: (a) assessment of missing values from nearby stations and (b) adjustment of the biases in the probability density function of the filled values towards the existing data of the target. The methodology is assessed for its performance in filling-in the time series of a dense precipitation station network with large gaps on the island of Crete, Greece. The results indicate that quantile mapping can benefit the filled-in missing data statistics, as well as the wet day fraction. Conceptual limitations of the method are discussed, and correct methodology application guidance is provided.

show abstract

Exploring the Impact of Various Spectral Indices on Land Cover Change Detection Using Change Vector Analysis: A Case Study of Crete Island, Greece

Cited by 60 publications

References 36 publications

Object-Oriented LULC Classification in Google Earth Engine Combining SNIC, GLCM, and Machine Learning Algorithms

Object-Oriented LULC Classification in Google Earth Engine Combining SNIC, GLCM, and Machine Learning Algorithms

Assessment of Intra-Annual and Inter-Annual Variabilities of Soil Erosion in Crete Island (Greece) by Incorporating the Dynamic “Nature” of R and C-Factors in RUSLE Modeling

A Quantile Mapping Method to Fill in Discontinued Daily Precipitation Time Series

Contact Info

Product

Resources

About