Field guide
Understand the geometry before saving a point
Keep the source observations, confirm the coordinate frame, and use independent evidence to choose a result.
Choose the construction that matches your observations
This COGO intersection calculator supports four horizontal constructions: direction–direction, direction–distance, distance–distance, and the intersection of two lines defined by four points. Choose the method from the evidence you actually have. A bearing bearing intersection needs an absolute direction at each known station. A bearing distance intersection combines a direction from A with a circle centered at B. A distance distance intersection uses two horizontal radii. Four-point mode defines A toward C and B toward D. These are coordinate geometry operations in a common plane; they do not perform a least-squares network adjustment or derive an instrument orientation from a backsight.
Keep the original observation record nearby. If your field book contains an included angle, a slope distance or an uncorrected ground measurement, prepare the appropriate direction and horizontal distance before entering it here. A numerical intersection can be internally consistent while representing the wrong physical construction. The written reference field is a useful place to record the project grid, drawing revision, units and reason for the calculation.
Keep east, north and direction conventions explicit
Enter Easting in E and Northing in N. Directions are clockwise from coordinate north: north is zero degrees, east is ninety, south is one hundred eighty and west is two hundred seventy. Choose decimal azimuth, spaced degrees minutes seconds, or a quadrant bearing before typing the value. For example, N 45 E is a quadrant direction; 45 30 00 in the azimuth DMS mode represents forty-five degrees and thirty minutes. A decimal point is not a substitute for the minute and second separators.
Use meters, international feet or US survey feet consistently. Changing the unit selector changes the declared unit; it does not transform any coordinate or radius. Latitude and longitude are unsuitable because degrees do not define a uniform horizontal distance in this calculation. Heights are deliberately excluded from the intersection. If you need an elevation for the saved point, enter a separately established value or leave that field blank.
Decide whether a bearing defines a ray or an infinite line
A forward ray begins at its known point and continues only in the stated direction. An infinite line extends both forward and backward. This distinction changes which solutions are admissible. A line may cross another line behind station A, yet that location cannot satisfy a forward observation from A. The tool filters such candidates when ray mode is selected and reports how many were excluded. It does not silently reverse the bearing to make a solution appear.
In four-point mode, A and C define one direction and B and D define the other. C and D establish orientation, not segment endpoints that limit the intersection. Choose rays when only forward continuation is meaningful, or lines when extensions on either side are allowed. A repeated coordinate pair cannot define a direction. Distinct parallel lines have no solution, while coincident lines can have infinitely many. Opposing collinear rays can overlap, meet only at their origins, or be disjoint.
Review both circle solutions before saving either one
Two circles commonly meet at two positions on opposite sides of the line connecting their centers. Both satisfy the entered distances. The first row is simply a stable display order; it has no authority as the intended field point. Use an independent direction, a site sketch, a known side of the baseline, or another checked observation to choose the correct candidate. The same issue occurs when an infinite line passes through a circle: two intersections may be valid until a ray condition removes one.
The candidate table and relative diagram show the known points and every retained solution together. Click a candidate row or its plotted marker to inspect and select it. Selection remains an explicit action even when there is only one solution. Saving is disabled until a candidate has been chosen. Enter a new point identifier that fits the destination job. The calculator does not check that identifier against a separate instrument database, so review naming conflicts before importing the file.
Treat weak geometry as a reason for another observation
Almost parallel directions can place an intersection very far from the known stations. A tiny bearing error may then cause a large coordinate change. The tool flags a small intersection-angle sine rather than claiming that a long displayed coordinate is reliable. The warning threshold is a computational review cue, not a survey accuracy specification. Your allowable uncertainty depends on the observation quality, control geometry, project requirements and how the derived point will be used.
Near tangency has a similar sensitivity. Two circle solutions can approach each other until they merge, and a small change in radius can remove both. Exact tangency is reported as one candidate with a warning. Separated circles, one circle strictly inside another, different concentric circles and identical circles are distinguished. No intersection is manufactured by taking the absolute value of a negative square root or adding an undocumented distance tolerance. Inputs outside the supported numerical range are blocked for review.
Read known points without losing their source
The optional file panel accepts local UTF-8 CSV or TXT data. Review the delimiter, header, column mapping and horizontal units before reading the records. You can then copy a known point into A, B, C or D by its physical source line. This avoids treating repeated point names as unique identifiers. Original fields and the source file remain in the JSON report, while the calculation uses the explicitly selected E and N values. A source record with an error cannot be used as a known point. XLSX workbooks are also accepted for this input: explicitly choose the worksheet, and replace formulas, dates or error cells with plain values before importing.
Manual editing after loading a known point clears that point’s file-selection label, so the report does not claim an unchanged source relationship. Any change to a calculation input invalidates the previous result and selection. Recalculate after changing a direction, radius, coordinate, unit or reference. The example is synthetic and is labeled accordingly. It demonstrates two solutions without implying that its coordinates came from an actual site survey.
Export a calculation record and a separate selected point
The all-candidates CSV is intended for review. The JSON report includes the entered conditions, result status, warnings, excluded-candidate count, source records and current selection. Save that report when you need to explain why one result was used and another was rejected. A separate PNEZD file contains only the selected new point, with Northing before Easting, the optional elevation and an intersection description. The selected-point file is raw point data; import text columns deliberately when opening it in spreadsheet software.
The screen uses six decimal places for convenient comparison, while reports retain original inputs and calculation precision. Derived decimal coordinate strings may contain additional digits produced by arithmetic; those digits are not evidence of field accuracy. Check the selected output by computing bearings and distances back to the known stations. Keep a copy of the original observations and the selection report alongside the final point file. Local processing means the point data is handled in this browser, without sending it to an AI service.