Remote sensing and landscapes
Sentinel-1/2, Landsat, MODIS, optical sensors, SAR and LiDAR; classification, time-series analysis and change detection.
Output: spatial and temporal baseline.A COMPLETE ANALYTICAL CHAIN
Observe, model, project and deploy.
My work starts with a specific territorial question. Earth observation characterizes landscapes and change; spatial ecology estimates distributions, habitat and connectivity; climate scenarios introduce future conditions; and geospatial engineering organizes the results into reproducible systems.
This continuity supports consistent spatial units, traceable assumptions, explicit validation and updateable information from initial processing through delivery to end users.
01Characterize the territory
02Model patterns and responses
03Evaluate future scenarios
04Operationalize results
FROM PIXEL TO LANDSCAPE
Repeated observation captures processes that one-off surveys miss.
I use optical and radar time series together with environmental data to classify land cover, detect disturbance and reconstruct snow, water, vegetation and land-use dynamics. These products become inputs to ecological models, climate analysis and operational monitoring.
INTEGRATED WORKFLOW
From heterogeneous data to a territorial analysis architecture.
Each stage solves a distinct part of the problem and produces inputs for the next: baselines, predictors, scenarios, models and decision layers.
Spatial ecology
Species records, survey effort, distribution models, habitat, connectivity and spatial validation.
Output: suitability, distribution and knowledge gaps.Climate scenarios
CMIP6, downscaling, bias correction, climate extremes and projected changes in exposure or habitat.
Output: comparable scenarios by time horizon.GeoAI and optimization
Machine learning, neural networks, ensembles, spatial optimization and network analysis for multivariate problems.
Output: predictions, priorities and critical relationships.Geospatial platforms
Spatial databases, APIs, cloud, Open Data Cube, STAC, Shiny and web applications for access and updating.
Output: operational geospatial intelligence.APPLIED CAPABILITIES
Solutions for environmental, territorial and operational questions.
Projects can range from a focused specialist analysis to a complete architecture covering processing, modelling, visualization and transfer to the client's technical team.
EO
Landscape classification and monitoring
For establishing baselines, detecting change and tracking environmental processes with consistent spatial coverage.
- Multi-temporal land-cover and land-use classification.
- Detection of disturbance, intervention, water, snow and vegetation change.
- Optical-radar integration for persistent cloud cover and complex signals.
- Data cubes and automated workflows for repeatable analysis.
Applications: territorial monitoring, biodiversity, restoration, operations and infrastructure.
SDM
Biodiversity, habitat and species
For turning scattered observations and environmental predictors into spatial evidence for monitoring and planning.
- Species distribution models and ensembles.
- Pseudo-absence design, bias treatment and spatial validation.
- Habitat modelling, connectivity and survey prioritization.
- Biodiversity assessment under present and future conditions.
Applications: conservation, environmental assessment, species monitoring and spatial prioritization.
CC
Climate scenarios and physical risk
For assessing how environmental conditions may change and what those changes imply for assets, ecosystems and territories.
- Selection and processing of CMIP6 models and scenarios.
- Downscaling, bias correction and climate-extreme indices.
- Hazard, exposure and vulnerability analysis.
- Time-horizon comparison, stress testing and adaptation prioritization.
Applications: spatial planning, infrastructure, mining, natural resources and adaptation.
AI
GeoAI, predictive models and analytical systems
For extracting signal from spatial and temporal datasets that require classification, prediction or multi-source integration.
- Machine learning and deep learning with Python and R.
- Feature engineering, validation and model optimization.
- Processing on AWS, Google Cloud and Google Earth Engine.
- Automated pipelines and deployment of analytical products.
Applications: advanced classification, spatial prediction, screening and decision-support systems.
CLIMATE + ECOLOGY + TERRITORY
Future scenarios change how the territory is interpreted.
I integrate climate projections with landscape, biodiversity and human-system information to assess changes in suitability, exposure, connectivity and spatial priority. The analysis can remain exploratory or progress to impact-specific and risk-specific modelling.
OPERATIONAL GEOSPATIAL INTELLIGENCE
Dashboards and platforms for querying models, scenarios and evidence.
The platform organizes the spatial evidence base, processing and model outputs in an environment that can be used by technical teams, managers and non-specialist users.
SPATIAL DBSCENARIOS
MODEL OUTPUTSDECISION LAYERS
01 · DATA
Consistent spatial evidence base
Raster, vector, observations, metadata and results under common rules for versioning, projection and traceability.
02 · PROCESSING
Reproducible pipelines
R, Python, SQL, cloud, containers and data cubes to update analyses without relying on manual steps.
03 · MODELS
Comparable scenarios
Results by horizon, scenario, species, hazard, asset or territory, with identifiable assumptions and versions.
04 · INTERFACE
Geoviewers and dashboards
Shiny and web applications for filtering, querying, comparing, visualizing and downloading results by user role.
05 · USE
Decision-oriented products
Hotspots, priorities, alerts, impact pathways, metrics and indicators linked to the project's core question.
APPLIED EXPERIENCE
The same technical foundation applied to different problems.
My experience combines scientific research, applied R&D, projects for mining and spatial planning, conservation, marine-coastal systems and geospatial data infrastructure.
Landscape classification and GeoAI
Machine-learning land-cover and land-use modelling in Fray Jorge Reserve; AI-based peatland classification in Magallanes; cloud geospatial workflows and work with Data Cube Chile.
Biodiversity, habitat and species
Geospatial assessment of biodiversity and critical habitat for mining in Chile and Canada; climate-informed habitat modelling; spatial modelling of pudu roadkill risk in Chiloé.
Climate, mountains and risk
Research on snow dynamics and rain-on-snow events in the Andes; climate-change analysis for spatial planning in Los Lagos; recent development of climate-risk methodologies and exploratory tools.
Environmental platforms and information systems
Integrated platforms for mangrove conservation and restoration in Colombia, development and research in Data Cube Chile, and participation in SIMA Austral, an operational information system for aquaculture.
Environmental Engineer, MSc in Environmental Resources, and PhD candidate in Biology and Applied Ecology. I research and develop Earth Observation and GeoAI solutions for climate change, macroecology, biodiversity, environmental monitoring and risk management.
My experience combines geospatial processing, predictive modelling, cloud infrastructure and information-system development, with applied work in Chile and projects developed in Canada, Colombia and other Latin American contexts.
Earth observationOptical, SAR, LiDAR, time series and classification.
Spatial ecologyHabitat, SDM, biodiversity, connectivity and prioritization.
Applied climate scienceScenarios, extremes, downscaling, hydrology and risk.
Geospatial engineeringPython, R, SQL, AWS, Google Cloud, ODC, GEE, STAC and Shiny.
COLLABORATION · CONSULTING · R&D
Projects that require spatial science and a usable implementation.
I can support projects from methodological design and proof of concept through processing, modelling and deployment of a geospatial platform for technical or operational use.
diego.ocampo.melgar@gmail.com