Six Ways to Choose Geospatial Survey Institute Data for Creating Heat Maps

When you think "I want to create a heat map," one of the first confusions is which Geospatial Survey Institute (GSI) data to use as the base. The publicly available information—maps, imagery, elevation, terrain classification, and so on—is broad, and if you choose based only on appearance, you may end up with a visually attractive map that is weak as a basis for judgment. In practice, the data you should choose depends on what you want to show with color, at what scale you will use it, and whether you need to confirm current conditions.

Table of contents

• Organize the types of GSI data before creating a heat map

• Choice 1: Decide on data by what you want to color-code

• Choice 2: Choose the format by whether you view areas or boundaries

• Choice 3: Decide resolution according to the level of detail needed

• Choice 4: Combine imagery-based data for current-condition confirmation

• Choice 5: Use terrain classification to give background meaning

• Choice 6: Check update timing and usage conditions before adopting

• Summary: If you choose GSI data correctly, heat maps become practical tools

Organize the types of GSI data before creating a heat map

The term "heat map" is convenient, but in reality it covers several creation methods. Some maps colorize elevation differences, some show point concentration with color, and some show attributes such as hazard spots or failure counts as areas. Therefore, the point to understand first is that "not all GSI data can be used directly as a heat map." The GSI publishes data that are easy to use as background maps, suitable for numerical analysis, useful as current-condition photographic confirmation, or helpful to read the genesis of the land; each has a different role.

For example, if you want to roughly grasp broad-area undulation, elevation-centered data such as color-elevation maps or elevation tiles are suitable. Elevation tiles are prepared with the same tile coordinates and pixel coordinates as map tiles and are available in text and PNG formats, making them convenient to overlay on maps. Conversely, if you want to see bias along roads or distributions around buildings, reference-line data such as the Fundamental Geospatial Data—which include road edges and building outer outlines—are useful. Moreover, if you want to confirm how the current ground surface appears, orthophotos are appropriate; if you want to explain why a bias appears in a location, land condition maps or terrain classification are effective. Thinking in terms of four roles—background, analysis, confirmation, interpretation—makes it easier to identify the data you need.

If you start work without clarifying this, you may load data heavier than necessary and slow processing, or you may overlook fine terrain differences because your data are too coarse. The practical point to grasp is to choose data that are neither insufficient nor excessive for what you need to judge, not to prioritize appearance. Understanding that creating a heat map begins with choosing a suitable base rather than simply coloring will significantly reduce rework later. Additionally, because GSI’s maps allow you to reference maps, imagery, elevation, terrain classification, and disaster information within the same interface, a practical workflow is often "first look to grasp the whole picture," then "select only the data needed for analysis." Rather than committing to a single dataset from the start, it is important to take an overview and then narrow down.

Choice 1: Decide on data by what you want to color-code

The first axis of decision is what you want to color-code. Without this, no matter how detailed the map you use, you won’t get a meaningful heat map. For example, if you want to see land elevation differences, slope steepness, or microtopographic differences in lowlands, elevation data should be central. Using elevation tiles or digital elevation models makes it easy to color by elevation values themselves and to treat undulation and relief quantitatively. Conversely, if you want to express concentration of people, facilities, or failures—i.e., "point clusters"—in gradations, you need to overlay your own data on background maps or boundary data.

Organizing this difference, GSI data can broadly be used in three ways. First is using elevation values themselves—appropriate for terrain heat maps. Second is using administrative boundaries, roads, buildings, etc., as positional references—effective for correctly placing distribution data on maps. Third is using terrain classification or land-condition data as background information—useful for explaining why color intensity appears in certain places.

A common practical mistake is deciding only to "make a heat map" without clarifying whether you want to view elevation, occurrence density, or historical comparison. For example, even within the same lowland, the needed data differ depending on whether you want to see elevation differences, susceptibility to inundation, or the history of embankment/cut-and-fill. For the former, elevation data; for the latter, terrain classification; for historical view, past imagery. If the subject you want to show with color changes, the primary dataset changes too. If you can state the "meaning of the color" in one sentence at the start, choosing GSI data becomes much clearer. In other words, decide the form of the question first, not the dataset name. The GSI data you use differ by question: "Where is low?", "Where is concentration?", "Where has it changed?", "Why is that a point of concern?" Simply following this order significantly reduces unnecessary downloads and re-layering.

Choice 2: Choose the format by whether you view areas or boundaries

Next, consider whether the object you want to view is an area or a line/boundary. A heat map generally evokes an image of area gradation, but in practice the optimal data format depends on whether you need an area background or lines/boundaries as references. If you want to view area-wide elevation differences or distribution tendencies, raster or mesh-type data are suitable. If you want to see bias along roads, rivers, or buildings, you need base lines or outlines.

The GSI Fundamental Geospatial Data include coastline, administrative boundaries and representative points, road edges, centerlines of railways, waterlines, and building outer polygons, making them convenient when you need to organize positional references. For example, if you place scattered inspection records or anomaly locations on a map and want to compare them with the placement of roads or facilities, having these basic line features helps prevent misreading where concentrations are occurring.

The important point here is not to confuse the heat map itself with background data. The background does not necessarily need to be flashy. In practice, the heat-map subject is the main element, so a background that makes positional relationships easy to read is more important. Whether you use fundamental data that clearly shows road and building outlines or a color-elevation map that shows topographic undulation will strongly affect how readable the finished figure is. If you want to show area changes but the boundaries are weak, pinpointing locations becomes difficult; conversely, if you want to show boundaries but the background is too complex, the color distribution may be obscured. Therefore, separating which format is the main focus and which format is supplementary is extremely important when choosing data. Also, when color-coding values aggregated by municipality or facility, how you present it depends on what you place underneath. If administrative boundaries matter, prioritize boundary lines; if you’re inspecting linear infrastructure like roads, prioritize road or waterline data so the viewer is not confused. A heat map may look like it stands on color alone, but in reality choosing the base that tells viewers "where to look" is crucial.

Choice 3: Decide resolution according to the level of detail needed

The third axis of decision is how finely you want to view the subject. GSI’s digital elevation models are available in multiple granularities such as 1 m (3.3 ft) mesh, 5 m (16.4 ft) mesh, and 10 m (32.8 ft) mesh. The 1 m (3.3 ft) is produced from airborne laser surveys, 5 m (16.4 ft) from airborne laser or photogrammetric surveys, and 10 m (32.8 ft) from contour lines on volcano base maps and topographic maps. Finer meshes capture more detailed relief but increase the data volume when the target area is large. Choosing a too-fine resolution for broad-area analysis increases processing load and can actually slow practical decision-making.

For example, for assessing trends at the municipal or watershed level, it is efficient to first grasp overall trends with somewhat coarser elevation data, then switch to higher-resolution data only in necessary areas. Conversely, when differences of several meters (several ft) matter—such as slight undulations in lowlands, terrace differences around residential development, or microtopography near roads—coarse elevation data cannot be read sufficiently. GSI’s maps include color-elevation maps and customizable elevation maps, so it is important to switch between broad-area understanding and microtopography readability.

Note that higher resolution is not always the right answer. A heat map only matters when viewers can judge differences. If you colorize excessively fine data as-is, local noise may be emphasized and it becomes hard to see what is truly important. Practically, what is needed is not the finest data but data that match the intended scale. Decide in advance whether the map is for showing broad trends in a meeting or for confirming local ground differences in the field; that makes it easier to judge whether to use 1 m (3.3 ft), 5 m (16.4 ft), or 10 m (32.8 ft). The same idea applies to field-confirmation materials: use coarse data to grasp wide-area trends, then increase detail only for the area where confirmation or countermeasures will be taken—this speeds the overall work. Do not aim to increase resolution itself; instead, determine the level of detail necessary for decision-making.

Choice 4: Combine imagery-based data for current-condition confirmation

Heat maps make numbers easy to see with color, but color alone sometimes cannot sufficiently explain field conditions. Combining imagery-based data is therefore important. GSI’s electronic national land basic map orthophotos convert aerial photographs into georeferenced images without positional displacement, so when you overlay elevation or distribution heat maps, it becomes easy to visually confirm relations with buildings, roads, development sites, farmland, and water bodies.

GSI has repeatedly compiled aerial photographs since the late 1940s, and they are available for viewing and download. Being able to refer to past imagery allows you to check whether current heat-map biases are influenced by historical terrain or land use. For example, a location that looks like residential land now might formerly have been a wetland or an old river channel. Judging based only on present color intensity without checking such history can lead to misinterpretation based on appearance alone.

In practice, it is very important to hold analysis data and confirmation data separately. Use elevation data or point-cloud–derived distribution data to find anomalies, then compare them to orthophotos or aerial photographs to confirm the current appearance. This workflow reduces the risk of ending with desk-based visualization only. The persuasive power of a heat map depends not on vivid color but on whether it aligns with current conditions. Therefore, when choosing GSI data, consider not only the main data for analysis but also the imagery-based data to overlay for confirmation. This is especially critical in areas where land-surface usage changes easily—such as developed sites, farmland converted lands, river corridors, and coastal zones—because judging by numbers alone without viewing imagery is risky. Whether the visible ground is paved or bare, whether buildings have increased, or whether waterways have been altered changes interpretation even for the same color. Placing orthophotos under a heat map greatly affects practicality.

Choice 5: Use terrain classification to give background meaning

The fifth axis is how to explain the meaning of the color. A heat map conveys intensity at a glance, but unless you explain "why it looks that way," it can remain a mere impression map. Land condition maps and vector-tile terrain classification are effective here. Land condition maps aim to provide baseline materials on natural land conditions necessary for disaster prevention planning, land use, land conservation, and regional development, generally showing terrain classifications such as mountainous areas, plateaus/terraces, lowlands, water areas, and artificial landforms.

Vector-tile terrain classification consolidates various terrain-classification datasets—such as land condition maps, flood-control terrain classification maps, and coastal area land condition maps—into a single layer stored as codes. This makes the background not just for visual inspection but also easy to handle by terrain type. For instance, an area that appears red for the same reason may lie along a natural levee, along a backswamp, or be concentrated in artificially altered land—interpretation differs by terrain class. Terrain classification fills context that color alone cannot convey.

Especially in operations such as disaster prevention, infrastructure maintenance, and land-use planning, showing the terrain context together with heat-map results makes judgment easier than showing the heat map alone. In practical explanations to superiors or stakeholders, saying "this place is a lowland" is vague, but saying "this lowland falls into such a terrain category, which makes this tendency likely" greatly increases the report’s persuasiveness. To give heat-map colors meaningful grounds for judgment, GSI terrain-classification data are very compatible. In regions where lowlands mix natural levees, backswamps, old channels, and artificially altered land, simple elevation differences alone often do not suffice for practical decisions. When a strongly colored area appears, whether you treat it as an elevation issue only or include land genesis affects the direction of measures. Therefore, when you want explanatory power in a heat map, include terrain-classification data as a candidate from the start rather than adding it later.

Choice 6: Check update timing and usage conditions before adopting

Finally, often overlooked are update timing and usage conditions. No matter how appropriate the data, ignoring creation dates or differences in geodetic results can cause mismatches in comparisons and overlays. GSI’s Fundamental Geospatial Data site handles past datasets for some basic items and certain digital elevation models, and update information for digital elevation models indicates that data maintenance and updates for 1 m (3.3 ft) and 5 m (16.4 ft) are ongoing. In recent years there have also been revisions to elevation results and updates to coordinate reference systems, so "using data you downloaded long ago as-is" is not always safe.

You should also confirm how to obtain the data. Some Fundamental Geospatial Data downloads require login, while others, such as GSI tiles, are convenient for real-time reading. If you create a heat map for public release, you must cite the source, and if you edit or process the data, you are required to indicate that. In other words, choosing data should include not only content but also how to obtain it, update frequency, and how to handle it when publishing.

In practice, first check the update timing for your target area, then align the coordinate system and vertical datum to be used, and finally confirm the rules for source citation to reduce errors. Especially when creating heat maps that compare multiple time periods, unconscious mixing of old photographs with new elevation data or new backgrounds with old boundaries makes it hard to tell whether a difference is a true change or a difference in data epochs. Before making polished figures, organizing when and under what standards the data were collected is indispensable for trustworthy heat-map creation. Also, as usage changes from internal use to reports, explanatory materials, or public pages, the checks required change. Practitioners tend to focus on map-making itself, but if there is a possibility of external submission or publication later, it is safer to assume citation and processing handling from the start. Checking usage conditions at the adoption stage rather than rushing checks right before publication stabilizes operations.

Summary: If you choose GSI data correctly, heat maps become practical tools

Choosing GSI data for heat-map creation looks difficult, but decisions become easier if you organize the sequence. First decide what you want to color-code. Next, distinguish whether you want to view areas or base your view on boundaries and lines. Then confirm necessary resolution, imagery for current-condition confirmation, terrain-based contextualization, and update timing and usage conditions—this naturally narrows the datasets to use.

The important point is not to try to complete everything with a single GSI dataset. In practice, combining different-role datasets—viewing terrain differences with elevation data, confirming current conditions with orthophotos, explaining background with terrain classification, and aligning positional relationships with Fundamental Geospatial Data as needed—produces more actionable heat maps. Prioritize making maps that are not misread in the field rather than merely attractive visuals; this keeps your data-selection criteria consistent.

Finally, even if you visualize results with GSI data, if you cannot confirm positions in the field, the work ends up as desk-bound material. If you want to check biases and focal points found in the heat map directly in the field and link them to simple surveying, combining with LRTK is very practical. As a smartphone-mounted high-precision GNSS positioning device, LRTK connects broad-area insights derived from public data with high-precision on-site position confirmation. Only then does a heat map become a living tool for decision and action rather than mere visualization.