Below you will find the definition of the terms commonly used in the description of the MapBiomas Venezuela methodology and products.

Word	Description
Algorithm	An algorithm is a finite sequence of rigorous instructions, typically used in mathematics and computer science to solve specific problems or perform computations. These instructions provide a step-by-step procedure to achieve a desired outcome. Algorithms serve as specifications for various calculations and data processing tasks.
Machine learning	Refers to the application of artificial intelligence techniques that enable computers to learn from and make predictions or decisions based on data. It involves training algorithms on remote sensing data to recognize patterns, classify objects, or extract valuable information automatically.
Empirical decision tree	In the context of machine learning and data analysis, decision trees are a popular algorithm used for classification and regression tasks.An empirical decision tree is a decision tree constructed based on observed data or empirical evidence.
Asset	A Google Earth Engine Asset refers to a piece of geospatial data or information stored and managed within the Google Earth Engine platform. They can include various types of geospatial data, such as satellite imagery, vector data and tables, among others. Google Earth Engine allows its users to upload, organize, and access assets for analysis and visualization purposes.
ATBD (Algorithm Theoretical Basis Document)	Is a comprehensive technical document that provides a detailed explanation of the theoretical foundations and principles behind a specific algorithm or data processing method. ATBDs are typically created in the context of Earth observation, remote sensing, and scientific research projects.
Band (or spectral band)	A spectral band represents a specific range of wavelengths within the electromagnetic spectrum detected by satellite-mounted sensors. These bands are crucial in remote sensing, as they enable the collection of data about the Earth's surface and its surroundings. They include visible bands for visible light, near-infrared bands beyond human vision, and thermal infrared bands for heat radiation. By analyzing data from different bands, remote sensing specialists can gain valuable insights on land cover, vegetation health, and temperature, among others, enhancing their understanding of the Earth.
Biome	A biome is a large-scale biological community characterized by distinctive plant and animal species adapted to a specific geographic area with similar environmental conditions. These conditions include factors such as climate, temperature, precipitation, and soil type. Biomes are major ecosystems that can cover extensive regions of the Earth and are often defined by their dominant vegetation types.
Random Forest	It is a machine learning algorithm used for image classification and data analysis. It belongs to the ensemble learning category, which means it combines the results of multiple machine learning models (decision trees) to make more accurate predictions or classifications.
Classification	It is the process of categorizing or labeling each pixel of an image into predefined classes or categories based on its spectral characteristics.
Classifier	It is a type of algorithm or model that is used to categorize or label data into predefined classes or categories. With satellite images the choice of classifier depends on the specific goals of the analysis, the complexity of the landscape, and the quality of available training data. Some commonly used classifiers include: Supervised Classification, Unsupervised Classification, Object-Based Classification, Decision Trees and Random Forest.
Collection	In MapBiomas, a collection refers to a specific version or dataset of their land cover and land use mapping project. Each collection represents a dataset covering a specific period and region. The collections are updated periodically, improving the spatial and/or thematic precision of the information and including new years, allowing users to access and analyze land cover changes over time up-to-date information.
RGB combination	In remote sensing, RGB combination refers to the process of creating a composite image using three different bands of data, typically representing the red, green, and blue wavelengths of light. Each band captures information from a specific portion of the electromagnetic spectrum. By combining these bands in an RGB composite, you can create a full-color image that is visually interpretable to the human eye.
Cloud computing	In Google Earth Engine, cloud computing refers to the use of remote servers hosted on the internet to store, manage, and process geospatial data and satellite imagery. This approach leverages the computational power and storage capacity of remote data centers, enabling users to access and analyze vast amounts of Earth observation data without the need for extensive local computing resources. The platform utilizes Google's cloud infrastructure to provide users with efficient and scalable tools for geospatial analysis, making it possible to perform tasks such as land cover classification, time series analysis, and change detection on a global scale. This cloud-based approach enhances accessibility and collaboration, making it a valuable resource for researchers, scientists, and environmental professionals in their geospatial analysis endeavors
Spatial consistency	In remote sensing, the term "spatial consistency" refers to the extent to which data captured by a sensor or satellite images maintain spatial coherence or a consistent spatial relationship in terms of patterns, distribution, and geospatial structures. Spatial consistency is essential to ensure that the data is reliable and suitable for use in remote sensing analysis and applications that require accurate geospatial information.
Temporal consistency	In remote sensing, the term "temporal consistency" refers to the stability and coherence of data collected over time by a sensor or observation platform, such as a satellite. It implies that the values of measurements or images captured at different time instances maintain a consistent and reliable relationship. Temporal consistency is crucial in remote sensing as it allows for precise tracking of changes and trends on the Earth's surface over time.
Dashboard	It is a component of the MapBiomas project that provides a visual interface in the Web, for accessing and exploring geospatial data related to land cover and land use. It serves as a user-friendly platform for viewing and analyzing information derived from remote sensing data and other sources. The Dashboard offers tools and interactive features that allow users to visualize changes in land cover over time, in reference to spatial units (states, watershed, protected areas, indigenous territories and many others), and access detailed data layers and statistics. It is a valuable resource for researchers, policymakers, and the general public interested in monitoring and understanding land cover dynamics and environmental changes
Ecoregion	An ecoregion is a distinct geographic area with specific ecological characteristics, such as climate, vegetation, and wildlife. These regions are defined based on environmental factors, serving as important units for biodiversity conservation and ecosystem management. They differ in scale and focus from biomes. An ecoregion is a smaller, geographically specific area with unique ecological characteristics, like climate and species. Biomes, on the other hand, are larger and encompass multiple ecoregions, characterized by broad vegetation and climate patterns. Ecoregions provide finer details for conservation and management, while biomes offer a broader overview of Earth's major ecological zones.
Code Editor	The Google Earth Engine Code Editor is an online development environment for creating and running geospatial analysis scripts. It provides access to a vast collection of Earth observation data and processing capabilities. Users can write JavaScript code to manipulate and analyze satellite imagery and geospatial datasets. The Code Editor offers a user-friendly interface for coding, visualization, and sharing of geospatial analyses.
Landsat scene	Refers to an individual image or snapshot of the Earth's surface captured by a Landsat satellite. A Landsat scene typically covers an area of approximately 170 kilometers (about 106 miles) in the North-South direction and 183 kilometers (about 114 miles) in the East-West direction.
Feature Space	En Google Earth Engine, el término “Feature space” (o espacio de características) se refiere a la representación de datos geoespaciales como entidades individuales. Las características son la unidad básica de datos en Earth Engine y pueden representar varios elementos geográficos, como puntos, líneas o polígonos. Cada característica suele contener tanto la geometría (ubicación) como las propiedades (atributos) que la describen. El espacio de características es fundamental para realizar análisis y manipulaciones geoespaciales dentro de Google Earth Engine.
Spatial filter	It is a technique used to enhance, modify, or analyze pixel values based on their spatial relationships within an image. These filters are commonly used in image processing and remote sensing to reduce noise, highlight certain features, or smooth data. For example, averaging filters calculate the average value of neighboring pixels to reduce noise, while edge-detection filters emphasize edges and boundaries within an image. Spatial filters play a crucial role in improving the quality and interpretability of classified images by refining pixel values according to their spatial context, making them valuable tools in image analysis and classification tasks.
Temporal filter	Is a computational technique used to enhance the quality of time-series data, particularly for remote sensing applications. It helps remove noise, artifacts, or inconsistencies present in multitemporal datasets acquired over time. Temporal filters can include methods like temporal smoothing or convolutional neural networks, which analyze data across multiple time steps to detect patterns, trends, and anomalies. By applying temporal filters, the accuracy and reliability of multitemporal classification, such as land cover change detection, are improved, making them valuable tools in remote sensing and environmental monitoring. These filters aid in extracting meaningful information from time-series data and reducing the impact of noise and errors.
Google Cloud Storage	Is a web service provided by Google Cloud Platform that offers RESTful online file storage. It allows users to store and access data in a scalable and secure manner using Google's infrastructure. Google Cloud Storage is designed to handle large volumes of data and is known for its performance and advanced security features. Users can leverage this service for various data storage needs, including backups, media content, and application data.
Google Earth Engine	It is a cloud-based geospatial platform by Google that provides access to an extensive archive of satellite and geospatial data. It allows users to analyze, visualize, and process this data for various applications, including environmental monitoring, land use analysis, and more.
Landsat image	A Landsat image is a digital photograph of the Earth's surface captured by a satellite within the Landsat program. These images have a moderate spatial resolution, meaning they provide a balance between capturing fine details and covering large areas. This makes them suitable for various land-cover and land-use studies. The images are composed of multiple spectral bands, each capturing different wavelengths of light. This multispectral data allows for the analysis of land features, vegetation health, water quality, and more.
Raster image	It i a type of digital image that represents data in a grid of pixels. Each pixel contains information about a specific location on the Earth's surface, making it a fundamental format for storing and analyzing geospatial data.
Spectral Index	In remote sensing, a spectral index is a numerical value calculated from the reflectance values of different wavelengths in the electromagnetic spectrum. These indices are used to quantify specific properties or features on the Earth's surface. For example, vegetation indices like NDVI (Normalized Difference Vegetation Index) assess the health and density of vegetation by comparing the reflectance in the red and near-infrared spectral bands. Spectral indices help remote sensing experts and scientists extract valuable information about land cover, vegetation health, water quality, and more from satellite or airborne imagery.
Visual interpretation	In remote sensing, visual interpretation refers to the process of manually analyzing satellite or aerial imagery to identify and classify features on the Earth's surface. It involves the human eye and expert knowledge to recognize and interpret objects, land cover, and changes over time. This method is essential for tasks such as land use mapping, disaster assessment, and environmental monitoring, where human expertise adds context and accuracy to the analysis.
Stable classes map	In MapBiomas, stable classes maps refer to thematic maps that represent pixels of land cover or land use classes that have remained unchanged or stable over a specific period of time. These classes typically include categories such as forests, water bodies, or permanent urban areas. Stable classes maps are important for monitoring and assessing long-term trends in land cover and land use, as they help identify areas that have maintained their characteristics and can serve as a baseline for comparison with changing land cover classes. These maps are valuable for environmental management, land use planning, and conservation efforts.
Integration Map	The term refers to the result of the process of combining or merging multiple thematic maps or layers of geospatial data into a single coherent spatial representation. These thematic maps may represent different aspects of land cover, land use, or landscape changes.
Transition Map	A transition map illustrates changes in land cover and land use over time. It shows how different land cover categories have transformed over the years. These maps provide crucial information about shifts in ecosystems, such as deforestation, urbanization, or agricultural expansion. By analyzing transition maps, researchers and policymakers can better understand the dynamics of land use changes and their environmental impacts, aiding in informed decision-making for sustainable land management and conservation efforts.
Land Use and Land Cover maps	They are graphical representations of the Earth's surface, showing different categories of both the physical covering of the land (land cover) and how humans use that land (land use). Land cover describes the types of natural and artificial features, such as forests and water bodies, while land use indicates the human activities taking place in those areas, like agriculture or industrial use. These maps are essential for environmental monitoring, urban planning, and natural resource management, providing valuable information for decision-makers and researchers.
Reference maps	In remote sensing, reference maps are detailed and accurate maps created using ground surveys, GPS, or high-resolution aerial imagery. They serve as a baseline for comparison with remote sensing data. These maps provide information about the Earth's surface, such as land cover types, infrastructure, and boundaries. Reference maps are essential for validating and calibrating remote sensing analyses, ensuring the accuracy of classification results, and assisting in applications like urban planning, environmental monitoring, and disaster management..
Image Mosaic	It is a composite image created by combining multiple individual images into a seamless and cohesive representation. This technique is often used to overcome limitations in spatial coverage, ensuring a broader area can be analyzed. Image mosaics are particularly valuable in remote sensing, where satellite or aerial imagery is stitched together to provide a comprehensive view of a large geographical area. Various algorithms and methods are employed to align and blend these images to create a continuous and detailed visual representation used for applications such as land cover mapping, environmental monitoring, and geographic information systems (GIS).
Training samples	Training samples are representative data points, polygons or pixels within an image that are used to teach a computer algorithm to recognize and classify specific features or land cover types. These samples are manually labeled to indicate the correct class or category they belong to, such as vegetation, water, or urban areas.
Precision samples	En teledetección, el término “muestras de precisión” generalmente se refiere a un conjunto de puntos de datos de verdad en el terreno o muestras de referencia que se utilizan para evaluar la precisión y exactitud de un algoritmo o modelo de clasificación. Estas muestras de precisión son esenciales para la validación y evaluación de la precisión en el contexto de la clasificación en teledetección.
Pixel	A pixel is the smallest element or unit in a digital image captured by remote sensing instruments, such as satellites or aerial sensors. Pixels are the building blocks of the image and are arranged in a grid pattern to create a complete picture of a given extent of the Earth's surface.
Post-classification	Refers to a process in which adjustments or refinements are made to a previously generated classification of remote sensing data. This involves correcting errors or reassigning categories after the initial classification. For example, if errors were identified in classifying a forest as barren land, corrections can be made to improve accuracy. Post-classification is important to ensure that the results accurately reflect reality and is commonly used in applications such as mapping and environmental monitoring.
Accuracy (Analysis)	Accuracy analysis in remote sensing is the process of evaluating the reliability and correctness of data obtained through remote sensing technologies. It assesses how well remote sensing data matches reality. This analysis involves comparing remote sensing data to ground-truth or reference data to measure the level of agreement or disagreement. Key factors in accuracy analysis include precision, bias, and errors in data acquisition, which can affect the quality and trustworthiness of remote sensing information. By conducting an accuracy analysis, remote sensing professionals can determine the limitations of their data and make informed decisions about its appropriate use in applications such as environmental monitoring, land use planning, and disaster management.
Spatial resolution	Spatial resolution refers to the level of detail or clarity in a satellite image. It represents the smallest discernible features on the Earth's surface that can be distinguished in the image. A higher spatial resolution means finer detail and the ability to distinguish smaller objects, while lower spatial resolution results in less detailed images with larger feature sizes. Spatial resolution is a crucial factor in remote sensing because it impacts the ability to identify and analyze specific land features, making it essential for various applications such as urban planning, agriculture, and environmental monitoring.
Scripts	In Google Earth Engine, a script is a piece of code written in JavaScript or Python that defines the processes for analyzing and visualizing geospatial data. These scripts can be used to access and manipulate satellite imagery, vector data, and more.
Satellite Sensor	A satellite sensor is an instrument or device mounted on a satellite that is designed to collect information about the Earth's surface and its environment. They capture various types of data, including imagery, radiation measurements, and other environmental parameters. The sensor operates by detecting and recording electromagnetic radiation, such as visible light, infrared, and microwave signals, that is either emitted or reflected from the Earth's surface. Los sensores del programa Landsat include a multispectral sensor that captures data in several spectral bands, including visible, near-infrared, and shortwave infrared (used to monitor land cover changes, vegetation health, and other Earth surface phenomena) and a Thermal Infrared Sensor designed to measure thermal radiation emitted from the Earth's surface.
Temporal series maps	Temporal series maps are a collection of maps derived from images captured over time, typically from satellites or aerial platforms, showing changes in the Earth's surface or features. They allow the observation of variations, such as seasonal changes in vegetation, urban growth, or natural disasters, by comparing multiple images acquired at different times. These maps are crucial for monitoring and analyzing long-term trends, enabling informed decision-making in fields like agriculture, environmental management, and disaster response.
Shapefile	A shapefile is a common geospatial vector data format used to store a variety of geographic vector features, including points, lines, and polygons. Shapefiles are widely employed for organizing and sharing geographic information.
Cross-cutting themes	In MapBiomas' classification methodology, "cross-cutting themes" refer to land cover or land use classes that are classified independently, without considering other classes simultaneously. This approach is used because, in general classifications involving multiple classes, these specific classes tend to get confused with others. Subsequently, an integration process is carried out between the overall classification and the cross-cutting themes to generate the final annual land cover and land use maps.
Workspace	In Google Earth Engine, it is a virtual environment where you can store and organize datasets, code scripts, and the results of your Earth Engine analyses. Workspaces are crucial for keeping your work organized and for collaborative efforts.