Sentinel-1 SAR flood mapping in Python: the 2022 Pakistan floods
How much land did the 2022 Sindh floods cover, and how fast did the water recede?
- 1,933 km² newly flooded of 8,095 km²
- Otsu threshold −14.1 dB
- 332 km² still flooded on 25 Dec
Satellite, LiDAR and map data, turned into answers you can check.
Geospatial consulting for teams who need flood extents, terrain models, water-quality trends or automated GIS workflows, delivered with the Python code that produced every number.
UK-based, working remotely. A written scope before any work starts. You keep the code.
Who this is for
- Environmental and water teams who need flood extents, reservoir or land-cover change measured from free satellite archives, repeatably.
- GIS teams with a backlog whose manual desktop steps should become a script that runs the same way every time.
- Consultancies and researchers who need a specialist pair of hands for SAR, LiDAR or terrain work inside a larger project.
- Product and data teams who want spatial data turned into a web map, dashboard or clean dataset others can use.
What we can do for you
-
Python geospatial automation and pipelines
Repeatable GeoPandas, Shapely and rasterio workflows that replace manual clicking: spatial joins, overlays, format conversion and scheduled processing, with spatial indexes so they stay fast at scale.
Fast vector geoprocessing in Python: spatial indexes, overlays and modern file formats -
Earth observation and remote sensing analysis
Sentinel-1 SAR flood mapping, Sentinel-2 water and vegetation indices, water-quality proxies and change detection, streamed from STAC catalogues so only the pixels you need are downloaded.
Sentinel-1 SAR flood mapping in Python: the 2022 Pakistan floods -
LiDAR, terrain and DEM processing
Point-cloud cleaning and classification, DTM, DSM and canopy height models, hillshade, slope, aspect, profiles and zonal statistics from national and global elevation data.
LiDAR point clouds in Python with laspy: from a LAZ file to a DTM, DSM and canopy height model -
Interactive web maps and dashboards
Shareable Leaflet maps built from your GeoDataFrames: styled layers, tooltips, popups and switchable basemaps, delivered as a single HTML file or embedded in your site.
Interactive web maps from Python with Folium and GeoPandas .explore() -
Data sourcing
Finding the right boundaries, elevation, imagery and statistics for a country or study area, checking licences and formats, and getting them into a usable, documented shape.
GIS data sources catalogue -
Training and workshops
Hands-on sessions for analysts moving from desktop GIS to Python, built around worked examples like the ones on this site and adapted to your own data.
Python geospatial tutorials -
3D-printed terrain models
Watertight, print-ready STL models of real landscapes from a DEM, scaled to your print size and checked before they reach the printer.
From DEM to 3D-printable STL with Python and NumPy
What this looks like
All tutorials →Four worked projects from this site. Each started with a question, used public data, and ends with figures you can reproduce from the published script.
Monitoring reservoir water quality from space with Sentinel-2 and Python (Peak District case study)
Can free imagery flag chlorophyll and turbidity changes in small upland reservoirs?
- 821 scenes → 122 usable dates
- Median NDCI 0.049 at Ladybower
- Checked against 18 EA samples
LiDAR point clouds in Python with laspy: from a LAZ file to a DTM, DSM and canopy height model
How tall are the trees on the National Mall, and how accurate is the surface model?
- 6.2 M points streamed in 15 s
- Monument measured to within 44 cm
- Median tree height 11.4 m
From DEM to 3D-printable STL with Python and NumPy
Can a free 30 m DEM become a physical model of the Edale valley?
- Copernicus 30 m DEM
- 150 mm watertight binary STL
- NumPy and rioxarray only
Check the work before you hire
The work is published in the open. There are 10 Python tutorials, 10 of them with their full scripts, requirements and recorded console output on a source code page. Each states the library versions it was tested with, and names every dataset and its licence.
The GIS data catalogue lists 880 sources for 237 countries and territories, link-checked automatically. It's the same groundwork that goes into sourcing data for a client project.
How an engagement works
-
Discovery call
A conversation about the question you need answered, the data you have, and what a useful result looks like for you.
-
Scoped proposal
A written scope: the deliverables, the data and methods, what is out of scope, the timeline and a quote.
-
Delivery
Work in agreed stages with early outputs to check against, so surprises in the data surface while there is still time to adjust.
-
Handover
Results, maps and the scripts that produced them, with notes on how to rerun them on new data. Nothing depends on us afterwards.
Ways to work together
Questions
Who owns the data, code and results?
You do. Everything produced for your project, including the scripts, is handed over at the end, and your own data is used only for your project.
What tools do you use?
Mainly Python: GeoPandas, Shapely, rasterio, rioxarray, pystac-client, laspy, OSMnx and Folium, plus QGIS and ArcGIS Pro where a desktop GIS fits better. If your team already works in a particular stack, the work can fit around it.
Do you use open-source software and open data?
By default, yes. Open libraries and free data such as Copernicus Sentinel, the Copernicus DEM, USGS 3DEP and OpenStreetMap keep costs down and mean you can rerun the work without licences. Commercial data or software can be used when a project needs it.
Will you sign an NDA?
Yes. Confidentiality terms can be agreed before you share any data or details of the project.
Where are you based, and do you work remotely?
TechnicalGIS is based in the UK and works remotely, so projects from clients elsewhere are welcome. The tutorials on this site were produced the same way, from public data and cloud-hosted archives.
How much does it cost?
Each project is scoped individually, so there is no price list. After a short discovery call you receive a written proposal with a quote before any work starts.
Tell us what you need to find out.
Describe the question, the area and any data you already have. You'll get a reply by email and, if it's a good fit, an invitation to a discovery call.
Start a consulting enquiry