Data sets for the e-sensing project

This project contains in situ data sets used in the e-sensing project. These data sets consist of time series from selected locations, which are used to train machine learning models, and data cubes to run examples of sits usage.

To load these datasets, first, install the sitsdata package using devtools:

devtools::install_github("e-sensing/sitsdata")

Next, load it:

library(sitsdata)

In the next sections are examples of how you can use the datasets available.

In case of any issue in following the steps above, please check the installation tip section.

Data format

The sitsdata R package contains time-series to be used for classification with machine learning methods which are available when the package is loaded using library(sitsdata). All satellite image time-series have the following columns:

• longitude (East-west coordinate of the time series sample in WGS 84).
• latitude (North-south coordinate of the time series sample in WGS 84).
• start_date (initial date of the time series).
• end_date (final date of the time series).
• label (the class label associated to the sample).
• cube (the name of the image data cube associated with the data).
• time_series (list with the values of the time series).

Datasets available

In the sections below is the metadata of each dataset available in the sitsdata package.

Land Use and Land Cover in Cerrado Biome using MODIS

The following table presents the metadata of this dataset:

Attribute	Details
Dataset ID	samples_cerrado_mod13q1
Region	Cerrado Biome (Brazil)
Number of Time Series	50160
Satellite	Terra
Satellite-Sensor	MODIS
Product	MOD13Q1
Data Source	NASA
Spatial Resolution	250 meters
Time Extent	2000-08-28 to 2019-09-01
Temporal Resolution	16-day composite (23 data points per year)
Spectral Bands	MIR, NIR
Spectral Indices	EVI,NDVI
Land Cover Classes	Dense_Woodland, Dunes, Fallow_Cotton, Millet_Cotton, Pasture, Rocky_Savanna, Savanna, Savanna_Parkland, Silviculture, Soy_Corn, Soy_Cotton, Soy_Fallow
Reference	Lorena Santos, Karine Ferreira, Gilberto Camara, Michelle Picoli, Rolf Simoes, “Controle de qualidade e redução de ruído de classe de séries temporais de imagens de satélite”. ISPRS Journal of Photogrammetry and Remote Sensing, vol. 177, pp 75-88, 2021. Access Link
License	Attribution 4.0 International (CC BY 4.0

To use this dataset, you can use the following command:

library(sits)
library(sitsdata)

data("samples_cerrado_mod13q1")

By using the command above, the dataset will be available in the samples_cerrado_mod13q1 variable.

Click to learn how to view the data

If you want to view the dataset you just loaded, you can use the sits R package:

plot(samples_cerrado_mod13q1)

To view it in am interactive map, use:

sits_view(samples_cerrado_mod13q1)

To learn more, please check the sits R package book.

Land Use and Land Cover in Mato Grosso using MODIS

The following table presents the metadata of this dataset:

Attribute	Details
Dataset ID	samples_matogrosso_mod13q1
Region	Mato Grosso State (Brazil)
Number of Time Series	1837
Satellite	Terra
Satellite-Sensor	MODIS
Product	MOD13Q1
Data Source	NASA
Spatial Resolution	250 meters
Time Extent	2000-09-13 to 2016-08-28
Temporal Resolution	16-day composite (23 data points per year)
Spectral Bands	MIR, NIR
Spectral Indices	EVI,NDVI
Land Cover Classes	Cerrado, Fallow_Cotton, Forest, Millet_Cotton, Pasture, Soy_Corn, Soy_Cotton, Soy_Fallow, Soy_Millet
Reference	Michelle Picoli, Gilberto Camara, et al., “Big Earth Observation Time Series Analysis for Monitoring Brazilian Agriculturee”. ISPRS Journal of Photogrammetry and Remote Sensing, 145: 328-339, 2018. Access Link Câmara, Gilberto; Picoli, Michelle, et al., “Land cover change maps for Mato Grosso State in Brazil: 2001-2017” (version 3). PANGAEA, 2021. Access Link
License	Attribution 4.0 International (CC BY 4.0

To use this dataset, you can use the following command:

library(sits)
library(sitsdata)

data("samples_matogrosso_mod13q1")

By using the command above, the dataset will be available in the samples_matogrosso_mod13q1 variable.

Click to learn how to view the data

If you want to view the dataset you just loaded, you can use the sits R package:

plot(samples_matogrosso_mod13q1)

To view it in am interactive map, use:

sits_view(samples_matogrosso_mod13q1)

To learn more, please check the sits R package book.

Land Use and Land Cover in a portion of the Cerrado Using CBERS-4 AWFI

The following table presents the metadata of this dataset:

Attribute	Details
Dataset ID	samples_cerrado_cbers
Region	Cerrado (Brazil)
Number of Time Series	922
Satellite	CBERS-4
Satellite-Sensor	MUX
Product	CB4_64_16D_STK
Data Source	BDC
Spatial Resolution	64 meters
Time Extent	2018-08-29 to 2019-08-13
Temporal Resolution	16-day composite (23 data points per year)
Spectral Bands	BAND13, BAND14, BAND15, BAND16
Spectral Indices	EVI,NDVI
Land Cover Classes	Cerrado, Fallow_Cotton, Forest, Millet_Cotton, Pasture, Soy_Corn, Soy_Cotton, Soy_Fallow, Soy_Millet
Reference	Karine Ferreira, Gilberto Queiroz, et al., “Earth Observation Data Cubes for Brazil: Requirements, Methodology and Products”. Remote Sensing, 12, 4033, 2020. Access Link
License	Attribution 4.0 International (CC BY 4.0

To use this dataset, you can use the following command:

library(sits)
library(sitsdata)

data("samples_cerrado_cbers")

By using the command above, the dataset will be available in the samples_cerrado_cbers variable.

Click to learn how to view the data

If you want to view the dataset you just loaded, you can use the sits R package:

plot(samples_cerrado_cbers)

To view it in am interactive map, use:

sits_view(samples_cerrado_cbers)

To learn more, please check the sits R package book.

Deforestation in Rondonia using Sentinel-2A

The following table presents the metadata of this dataset:

Attribute	Details
Dataset ID	samples_prodes_4classes
Region	Rondonia (Brazil)
Number of Time Series	393
Satellite	Sentinel-2
Satellite-Sensor	MSI
Product	SENTINEL-2-L2A
Data Source	BDC
Spatial Resolution	10 meters
Time Extent	2020-06-04 to 2021-08-26
Temporal Resolution	16-day composite (29 data points per year)
Spectral Bands	B02, B03, B04, B08, B8A, B11, B12
Spectral Indices	NDVI, EVI, NBR
Land Cover Classes	Burned_Area, Forest, Highly_Degraded, Cleared_Area
Reference	SITS Access Link
License	Attribution 4.0 International (CC BY 4.0

To use this dataset, you can use the following command:

library(sits)
library(sitsdata)

data("samples_prodes_4classes")

By using the command above, the dataset will be available in the samples_prodes_4classes variable.

Click to learn how to view the data

If you want to view the dataset you just loaded, you can use the sits R package:

plot(samples_prodes_4classes)

To view it in am interactive map, use:

sits_view(samples_prodes_4classes)

To learn more, please check the sits R package book.

Yearly deforestation samples in the Amazon Biome using Sentinel-2/2A

The following table presents the metadata of this dataset:

Attribute	Details
Dataset ID	samples_deforestation
Region	Amazon Biome (Brazil)
Number of Time Series	6007
Satellite	Sentinel-2
Satellite-Sensor	MSI
Product	SENTINEL-2-L2A
Data Source	BDC
Spatial Resolution	10 meters
Time Extent	2022-01-05 to 2022-12-23
Temporal Resolution	16-day composite (23 data points per year)
Spectral Bands	B02, B8A, B11
Spectral Indices	NDVI, EVI, NBR
Land Cover Classes	Clear_Cut_Bare_Soil, Clear_Cut_Burned_Area, Clear_Cut_Vegetation, Forest, Mountainside_Forest, Riparian_Forest, Seasonally_Flooded, Water, Wetland
Reference	SITS Access Link
License	Attribution 4.0 International (CC BY 4.0

To use this dataset, you can use the following command:

library(sits)
library(sitsdata)

data("samples_deforestation")

By using the command above, the dataset will be available in the samples_deforestation variable.

Click to learn how to view the data

If you want to view the dataset you just loaded, you can use the sits R package:

plot(samples_deforestation)

To view it in am interactive map, use:

sits_view(samples_deforestation)

To learn more, please check the sits R package book.

Image Data cubes used for classification examples

Indices and spectral bands from SENTINEL-2/2A images for tile 20LMR in Rondonia

The following table presents the metadata of this dataset:

Attribute	Details
Dataset ID	Rondonia-20LMR
Region	Rondonia (Brazil)
Number of Images	23 TIF files
Satellite	Sentinel-2
Satellite-Sensor	MSI
Spatial Resolution	10 meters
Time Extent	2022-01-05 to 2022-12-23
Tile	20LMR (MGRS grid)
Temporal Resolution	16-day composite
Spectral Bands	B02, B8A, B11
Spectral Indices	NDVI, EVI, NBR
Reference	SITS Access Link
License	Attribution 4.0 International (CC BY 4.0

To use this dataset, you can use the following command:

library(sits)
library(sitsdata)

data_dir <- system.file("extdata/Rondonia-20LMR", package = "sitsdata")

By using the command above, the path to the Rondonia-20LMR dataset will be stored in the data_dir variable.

To learn more, please check the sits R package book.

EVI and NDVI MOD13Q1 images for Sinop

The following table presents the metadata of this dataset:

Attribute	Details
Dataset ID	sinop
Region	Sinop (Brazil)
Number of Images	23 TIF files
Satellite	Terra
Satellite-Sensor	MODIS
Spatial Resolution	250 meters
Time Extent	2013-09-14 to 2014-08-29
Tile	012010
Temporal Resolution	16-day composite
Spectral Indices	NDVI, EVI
Reference	SITS Access Link
License	Attribution 4.0 International (CC BY 4.0

To use this dataset, you can use the following command:

library(sits)
library(sitsdata)

data_dir <- system.file("extdata/sinop", package = "sitsdata")

By using the command above, the path to the sinop dataset will be stored in the data_dir variable.

To learn more, please check the sits R package book.

Installation tips

If you are having network issues installing the sitsdata package, please consider increasing the timeout limit:

options(timeout = 300) # Set timeout to 5 minutes

After updating the timeout limit, you can install the package using the devtools command:

devtools::install_github("e-sensing/sitsdata")

If you have any other issues, please ask for help in the issue section; we are keen to support you!