How We Actually Draw Site Analysis Diagrams
Most people treat site analysis diagrams as a checklist exercise. They trace property lines, mark north, stick an arrow pointing at the sun, maybe color in a wind rose, and call it done. That's not how they work in practice. A real diagram is shorthand for decisions that will mess with your program later. I learned that the hard way on a small residential project a few years back. The site was sloped, bordered by a highway on one side and a dense tree line on the other. I spent two days drawing sun paths, wind vectors, noise contours, sightlines, and accessibility routes. Got it looking clean. Then the client showed up with a completely different set of priorities than what the brief said, and half my diagrams were suddenly irrelevant. The site analysis didn't anchor to the actual design problem. It just sat there as decoration. Since then I've stopped treating it as a standalone deliverable. Now I tie every diagram back to a design decision before I even open the CAD file. If a vector doesn't influence wall placement, window size, massing, or entry, I cut it.
What Architectural Site Analysis Diagrams Are For
They are visual summaries of environmental, contextual, regulatory, and topographic data that directly affect building placement and orientation. That's the definition you'll find in textbooks. The part nobody writes down is that they only matter when you use them during the conceptual phase, not after. Beginners often produce diagrams that read like maps rather than arguments. A map shows where things are. An analysis diagram shows why the building should go somewhere specific. Those are different things.
The Process I Use
Step one is always data gathering, but not in the way most people do it. I don't start with satellite imagery. I start with constraints. Setbacks, height limits, easements, flood zones, utility connections, road access points, neighboring building heights, dominant sightlines, existing vegetation worth keeping. Whatever the municipality makes non-negotiable. I pull this from public GIS portals, municipal zoning maps, and utility records. Sometimes I have to visit the site. Not for photos. For orientation. Standing on a lot tells you something flat maps cannot. I once spent twenty minutes just watching how wind moved through a narrow corridor between two mid-rise buildings. That informed a ventilation strategy more than any weather station data. Step two is organizing that data into layers. I keep it minimal. Topography, solar path, wind, noise, views, circulation, and zoning. Any more layers and the diagram becomes unreadable. Any fewer and you miss the conflict points.
Get the Full Details
Step three is making the diagrams themselves. I usually work in Rhino or SketchUp for massing studies, then bring everything into Revit or Illustrator for final presentation. AutoCAD still works fine if that's your workflow. I don't believe in tool loyalty here. The diagram quality depends on clarity, not software. Solar analysis takes the longest. I use shadow studies at key dates: winter solstice, equinox, summer solstice. The diagram itself is simple. Sun arcs, shadow polygons at morning, noon, and afternoon, and a heat map showing which parts of the site receive direct light versus shade throughout the year. What beginners miss is that you need to overlay this onto your proposed massing, not just the empty lot. An empty lot's solar pattern is academic. Your building's shadow pattern is what matters. Wind diagrams are usually drawn as vector arrows showing dominant wind direction and velocity zones across the site. I use basic CFD if the project is large enough to justify it. For smaller projects, I approximate using local weather data and site shape. Again, overlay this on the massing.
Noise and circulation diagrams are often where site analysis falls apart. I draw noise contours from adjacent roads, railways, or industrial sites. Then I draw pedestrian and vehicular circulation paths. The trick is finding where they conflict. That's usually where the building should not go.
A Specific Problem I Ran Into
There was a commercial project on a corner lot in a mixed-use district. The site analysis showed clear solar and wind patterns. Everything looked fine on paper. Then I tried to integrate the accessibility routes with the fire lane requirements, and the two systems clashed completely. The fire lane needed a ten-meter-wide continuous loop. The accessible route needed a maximum gradient of five percent. On that particular topography, satisfying both meant pushing the building massing twenty meters further from the street than the zoning allowed. The workaround was surprisingly simple. I dropped the fire lane slightly below grade in one section, creating a shallow ramp that also served as the accessible route. Combined both diagrams into a single layered drawing instead of keeping them separate. Saved the design. Took about three hours of redoing the diagrams, which is still faster than redesigning after permit review. This is the part about site analysis nobody highlights: the diagrams force conflicts to surface early. If you wait until schematic design to resolve them, the cost of changes multiplies.
Common Pitfalls
First pitfall is producing too much data without synthesis. A site analysis diagram with fifteen overlapping layers looks thorough. It is not. It is illegible. Pick the four or five that drive the biggest decisions and show those clearly. Use color sparingly. Blue for water features, red for solar exposure, green for vegetation, gray for built context. Stick to a palette and don't expand it. Second pitfall is ignoring microclimate. Textbooks talk about climate zones. Real sites have microclimates. A shaded courtyard next to a reflective glass facade will behave differently than the same courtyard two blocks away. Measure what's there, not what the nearest weather station says. Third pitfall is treating the diagram as finished when it isn't. Site analysis is iterative. Every time you move a wall or shift a program block, the diagram needs to be checked again. I update mine at each major design milestone. Sometimes that means a five-minute refresh. Sometimes it means redoing the whole set. Either way, an outdated site analysis is worse than none at all because it creates false confidence.
Another issue people overlook is seasonal variation. Some sites behave completely differently in winter versus summer. Snow accumulation patterns, seasonal flooding, leaf canopy in summer versus bare branches in winter. I always include at least two seasonal snapshots. It adds maybe twenty percent to the diagram time but prevents embarrassing mistakes later.
When This Method Fails
Site analysis diagrams are not useful on flat, featureless sites with no regulatory constraints and no adjacent buildings. If the lot is a blank slate in an industrial park with no environmental factors and no context, the diagram will be empty and the exercise is wasted time. In those cases, a simple bubble diagram with program adjacencies does the same job faster. They also fail when the site data is incomplete or inaccurate. Municipal GIS maps can be outdated. Property lines can be wrong. Easement records can be missing. I've seen entire site analyses discarded because a buried utility easement appeared six months into construction. Always verify critical data with a physical survey if the project budget allows it. A professional land survey costs money but prevents far more expensive surprises.
What to Include in Each Diagram
Solar diagram: sun arcs for solstice and equinox, shadow polygons at three time points per date, shading ratio across the site, and the proposed massing shadow on the worst-case date. Wind diagram: dominant wind directions with velocity indicators, wind channeling effects between buildings, sheltered zones, and the building footprint exposed to prevailing winds. Noise diagram: decibel contours from identified sources, quiet zones, and the building areas most affected.
Circulation diagram: pedestrian paths, vehicle access, emergency access, bike routes, transit proximity, and conflict points between them. Topography diagram: contour lines at one-meter intervals minimum, slope percentage zones, drainage patterns, and areas prone to pooling or erosion. Context diagram: neighboring building heights, uses, architectural character, significant views to preserve or block, and vegetation to retain.
Zoning diagram: setbacks, height restrictions, floor area ratio, permitted uses, parking requirements, and green space mandates. Overlap this with the site constraints layer.
Final Notes
The diagrams are only as good as the questions they answer. Before you draw anything, write down what design decisions the site analysis is supposed to inform. If you can't connect a diagram to a specific decision about orientation, massing, entry, or program placement, remove it. Clutter in a site analysis is the same as clutter in a floor plan. It just looks professional while doing nothing useful. I keep a standard template set for recurring project types. Residential gets a different emphasis than commercial. A hotel site needs privacy and view analysis that a warehouse never will. Don't copy-paste diagrams across project types. The data changes enough that a reused diagram is usually lying to you. If you want something to work from, most major architecture firms maintain internal templates. You can find similar sets shared on sites like ArchDaily or BIMobject, but adapt them. A template is a starting point, not a substitute for thinking about your particular site.
