GeographicLib.NET

GeographicLib is a small set of C++ classes for performing conversions between geographic, UTM, UPS, MGRS, geocentric, and local cartesian coordinates,for gravity (e.g., EGM2008), geoid height and geomagnetic field (e.g., WMM2020) calculations, and for solving geodesic problems.

GeographicLib.NET is a native .NET implementation of GeographicLib written in pure C#.

What's different from NETGeographicLib

Unlike NETGeographicLib, GeographicLib.NET is implemented in pure C# without binding the C++ GeograpbicLib library by using C++/CLI or P/Invoke, thus achieves higher level of portability.

You should be able to use GeographicLib.NET with any target framework and platform that supports .NET Standard 2.0 or above.

Features

Bellow is a list of implemented features.

• Projections (AlbersEqualArea, AzimuthalEquidistant, CassiniSoldner, Gnomonic, LambertConformalConic, PolarStereographic and TransverseMercator)
• Geocodes (GARS, Geohash, Georef, MGRS and OSGB)
• Coordinate conversions (UTMUPS, Geocentric and LocalCartesian)
• Coordinate parsing/formatting (DMS and GeoCoords)
• Geodesic (Geodesic, GeodesicLine, GeodesicExact and GeodesicLineExact)
• Geodesic intersections (Intersect)
• Rhumb (Rhumb and RhumbLine)
• PolygonArea (PolygonArea<T>, PolygonArea, PolygonAreaExact and PolygonAreaRhumb)
• Geoid (Geoid)
• GravityModel (GravityCircle, NormalGravity and GravityModel)
• MagneticModel (MagneticModel, MagneticCircle)
• Auxiliary classes (MathEx, Ellipoid, EllipticFunction, SphericalHarmonic, AuxAngle, AuxLatitude and DAuxLatitude)

Geodesic and GeodesicExact are tested with the test set for geodesic.

TransverseMercator and TransverseMercatorExact are tested with data generated by 64-bit TransverseMercatorProj utility ran on Windows.

Managed implemetation of C mathematical functions in MathEx are tested with data generated by 64-bit Windows Universal C Runtime.

Installing

Stable

Stable releases of GeographicLib.NET are hosted on NuGet. You can install them using the following command:

dotnet add package GeographicLib.NET

Preview

Preview versions of GeographicLib.NET are hosted on NuGet pre-release channel. You can install them using the following command:

dotnet add package GeographicLib.NET --prerelease

Usage

This section lists some common usage examples of GeographicLib.NET. For detailed documentation, please refer GeographicLib's documentation.

Inverse calculation

Distance from JFK to LHR.

double
  lat1 = 40.6, lon1 = -73.8, // JFK Airport
  lat2 = 51.6, lon2 = -0.5;  // LHR Airport

double arcLength = Geodesic.WGS84.Inverse(lat1, lon1, lat2, lon2, out double distance);

Direct calculation

The point 5500 km NE of JFK:

double
  lat1 = 40.6, lon1 = -73.8,
  s12 = 5500e3, azi1 = 51;

double arcLength = Geodesic.WGS84.Direct(lat1, lon1, azi1, s12, out double lat2, out double lon2);

Geodesic intersection

Find closest intersection of two geodesics:

Geodesic geod = Geodesic.WGS84;
Intersect inter = new Intersect(geod);
IGeodesicLine
    lineX = geod.Line(0, 0, 45),
    lineY = geod.Line(45, 10, 135);

// Find displacement to closest intersection.
// Where point.X is the displacement from the starting point of lineX,
// and point.Y is the displacement from the starting point of lineY.
Point point = inter.Closest(lineX, lineY);

// Get the position of the intersection point by using the displacement on lineX.
lineX.Position(point.X, out double latx, out double lonx);

// Get the position of the intersection point by using the displacement on lineY.
lineY.Position(point.Y, out double laty, out double lony);

// (latx, lonx) and (laty, lony) should be pointing to the same location.
Assert.AreEqual(latx, laty, 1e-12);
Assert.AreEqual(lonx, lony, 1e-12);

Mathematical Functions

GeographicLib uses several C mathematical functions that are not available in all versions of .NET. These functions are:

• remquo
• frexp
• sincos (available since .NET 7.0, but not used due to this issue)
• hypot (available since .NET 7.0, but not used due to this issue)
• log1p (a naive implementation is available since .NET 7.0, not used by GeographicLib.NET)
• expm1 (a naive implementation is available since .NET 7.0, not used by GeographicLib.NET)
• exp2 (a naive implementation is available since .NET 7.0, not used by GeographicLib.NET)
• log2 (available since .NET 5.0)
• fma (available since .NET 5.0)
• scalbn (available since .NET 5.0)
• copysign (available since .NET 5.0)
• atanh (available since .NET Standard 2.1)
• asinh (available since .NET Standard 2.1)
• cbrt (available since .NET Standard 2.1)

GeographicLib.NET provides managed implementations (ported from musl libc) and platform dependent native C wrappers for these functions.

By default, the library will use the managed implementation when the corresponding math function is not provided by .NET runtime. You can also force the library to use platform dependent native C wrappers, by setting GeographicLib.MathEx.UseManagedCMath property to false.

Performance

BenchmarkDotNet v0.13.12, Windows 11 (10.0.22621.3155/22H2/2022Update/SunValley2)
Intel Xeon CPU E5-2689 0 2.60GHz, 1 CPU, 16 logical and 8 physical cores
.NET SDK 8.0.200
  [Host]   : .NET 6.0.27 (6.0.2724.6912), X64 RyuJIT AVX
  .NET 6.0 : .NET 6.0.27 (6.0.2724.6912), X64 RyuJIT AVX
  .NET 7.0 : .NET 7.0.16 (7.0.1624.6629), X64 RyuJIT AVX
  .NET 8.0 : .NET 8.0.2 (8.0.224.6711), X64 RyuJIT AVX

Method	Job	Runtime	Target	Mean	Error	StdDev	Ratio	RatioSD	Allocated	Alloc Ratio
Direct	.NET 6.0	.NET 6.0	Geodesic	814.7 ns	15.27 ns	17.59 ns	1.00	0.00	-	NA
Direct	.NET 7.0	.NET 7.0	Geodesic	765.3 ns	10.03 ns	9.38 ns	0.94	0.02	-	NA
Direct	.NET 8.0	.NET 8.0	Geodesic	804.1 ns	12.70 ns	14.63 ns	0.99	0.02	-	NA
Inverse	.NET 6.0	.NET 6.0	Geodesic	2,440.2 ns	24.36 ns	20.34 ns	1.00	0.00	-	NA
Inverse	.NET 7.0	.NET 7.0	Geodesic	2,278.6 ns	25.80 ns	22.87 ns	0.93	0.01	-	NA
Inverse	.NET 8.0	.NET 8.0	Geodesic	2,084.5 ns	18.19 ns	16.12 ns	0.85	0.01	-	NA
Direct	.NET 6.0	.NET 6.0	GeodesicExact	2,414.8 ns	44.92 ns	44.12 ns	1.00	0.00	-	NA
Direct	.NET 7.0	.NET 7.0	GeodesicExact	2,335.3 ns	19.53 ns	17.31 ns	0.97	0.02	-	NA
Direct	.NET 8.0	.NET 8.0	GeodesicExact	2,331.7 ns	18.35 ns	17.17 ns	0.97	0.02	-	NA
Inverse	.NET 6.0	.NET 6.0	GeodesicExact	6,057.7 ns	50.90 ns	47.61 ns	1.00	0.00	-	NA
Inverse	.NET 7.0	.NET 7.0	GeodesicExact	6,029.2 ns	51.33 ns	48.02 ns	1.00	0.01	-	NA
Inverse	.NET 8.0	.NET 8.0	GeodesicExact	5,947.1 ns	31.88 ns	28.26 ns	0.98	0.01	-	NA
Direct	.NET 6.0	.NET 6.0	Rhumb	1,057.2 ns	12.27 ns	11.47 ns	1.00	0.00	-	NA
Direct	.NET 7.0	.NET 7.0	Rhumb	1,046.5 ns	16.96 ns	15.04 ns	0.99	0.02	-	NA
Direct	.NET 8.0	.NET 8.0	Rhumb	943.8 ns	7.27 ns	6.07 ns	0.89	0.01	-	NA
Inverse	.NET 6.0	.NET 6.0	Rhumb	846.6 ns	6.29 ns	5.89 ns	1.00	0.00	-	NA
Inverse	.NET 7.0	.NET 7.0	Rhumb	841.4 ns	8.24 ns	7.30 ns	0.99	0.01	-	NA
Inverse	.NET 8.0	.NET 8.0	Rhumb	757.8 ns	6.67 ns	5.91 ns	0.90	0.01	-	NA

Change Log

GeographicLib.NET adopts changes made in GeographicLib and aligns its version number with GeographicLib releases.

Bellow is a list of stable releases of GeographicLib.NET and changes made in .NET side in each release. For changes adopted from GeographicLib, please refer the its change log here.

Version 2.3.2 (released 2024/10/23)

• FIX
  • Strong name validation fails when loading GeographicLib.dll from .NET Framework application

Version 2.3.1 (released 2024/03/20)

• NEW
  • Improve performance of Geodesic.Direct, Rhumb.Direct, GeodesicExact.Direct, GeodesicExact.Inverse and their overloads/variants. These are now heap allocation free.
  • Allow constructing MagneticModel and GravityModel from Stream and byte array. (#30)
  • AuxLatitude now implements IEllipsoid interface.
  • Parameter p0 of Intersect.Closest() now defaults to Point.Zero.
• FIX
  • Fixed wrong GravityFlags handling in GravityCircle class.

Version 2.3.0 (released 2024/02/23)

• NEW
  • Target .NET 7.0 and .NET 8.0.
  • Add managed implementation and libc wrapper for exp2.
  • Managed implementations and libc wrappers for log1p, expm1 and exp2 are now used for all versions of .NET, because the implementations provided by .NET runtime are just wrappers over Math.Log and Math.Pow.
• FIX
  • PolygonAreaExact seems to give incorrect results.

Version 2.1.2 (released 2023/01/17)

• FIX
  • Fixed an issue where number and date time parsing may fail or produce unexpected result when working with number and date time formats provided by CultureInfo.CurrentCulture. This fixes issue #27.

Version 2.0.0 (released 2022/08/24)

• NEW
  • Add IPolygonArea interface to provide better support for unit testing and dependency injection.
  • Add Count and IsPolyline properties (corresponding to NumPoints and Polyline methods in GeographicLib) to PolygonArea.
  • Add a new method Utility.FractionalYear which can parse floating point number and date time string as fractional year.
  • Allow constructing Geoid from Stream and byte array.
• FIX
  • String accepting APIs now handle lowercase "nan" correctly.
  • Ensure consistent rounding mode (MidpointRounding.ToEven) across different runtimes when converting floating point number to string.
  • MGRS.Reverse now allows lowercase input.

Version 1.52.1 (released 2022/04/12)

• NEW

  • Target .NET 6.0 in addition to .NET 5.0, .NET Standard 2.1 and .NET Standard 2.0.
  • Source Link support.

• BREAKING

  • Fixed typos in Ellipsoid. (Renamed SecondFlatterning to SecondFlattening and ThirdFlatterning to ThirdFlattening)

• FIX

  • Fixed an issue that Freeze method in AlbersEqualArea, LambertConformalConic and PolarStereographic was not working correctly.
  • Fixed duplicate instantiation of WGS84 and GRS80 static properties defined in NormalGravity.
  • Add missing support for World Magnetic Model Format v2.

Version 1.52.0 (released 2021/07/07)

• BREAKING

  • Modify overloads of Forward and Reverse in AlbersEqualArea, AzimuthalEquidistant, CassiniSoldner and LambertConformalConic to return coordinates as tuples.
  • Modify methods using out parameters defined in NormalGravity, GravityModel, GravityCircle, MagneticModel and MagnegticCircle to return results as tuples.

• NEW

  • Add constructor overloads that accept IEllipsoid as parameter for AlbersEqualArea and LambertConformalConic.
  • Add managed implementation of log2.
  • Add overloads of Direct and Inverse in Geodesic/GeodesicLine, GeodesicExact/GeodesicLineExact and Rhumb/RhumbLineExact, that return the computation results as a single object.
  • Add definitions of popular reference ellipsoids in Ellipsoid class.

• FIX

  • Fix stack overflow bug for Forward(double lon0, double lat, double lon) and Reverse(double lon0, double x, double y) in TransverseMercatorExact.

Version 1.51.0 (released 2021/03/14)

Initial stable release.