2020年3月9日发布的GAMIT10.71已经支持北斗三代数据解算；官方给出的说明如下：

- Beidou processing now uses L6/C6 (B3 at 1268.53 MHz) rather than L7/C7 (B2b at 1207.140 MHz) as the lower-frequency observable to accommodate the tracking configuration of most of the current GNSS receivers providing public data.Based on a sampling of 230 RINEX files from the archives, about 50% include only L6/C6 compared with about 5% that include only L7/C7. The higher-frequency observable will remain L2/C2 (B1-2 at 1561.098 MHz) since very few of the public RINEX files contain L1/C1 from the B1 at 1575.42 MHz transmitted by BDS 3 satellites. If L7/C7 is the only observak present (i.e, no L6/C6),GAMIT will try to process the data along with L7/C7 receivers but there may be ambiguity resoluion issues. User should data files and process only those sites with the same frequency data.  There are still some issues with clock estimates for Beidou data and pre-fit edting in autlcn.

查看了源代码（gamit/lib/sel_obtyp.f），GAMIT目前支持北斗观测值如下：

本文测试GAMIT处理北斗三代数据，利用LHAZ和URUM测站2020年4月1日的长基线解算，遇到如下问题：

1）sh_sp3fit -f gbm20993.sp3 -o gfzc -d 2020 092 -gnss C

STATUS :200427:1114:31.0 ARC/arc:   Integrating satellite 34  C38  220 3I       PRN 38

WARNING:200427:1114:31.0 ARC/ertorb: Radiation scaling unknown for BEIDOU-3I       : use GPS Block IIF value

 No MASS for BLKNUM            0

STATUS :200427:1114:32.0 ORBFIT/orbits/orbfit: Started ORBFIT ver. 10.15 2020/02/3 17:00:00 UTC (Linux) Library ver. 11.33 of 2020/02/10 14:24 UTC (Linux)

FATAL  :200427:1114:32.0 ORBFIT/lib/topens: Error reading 1st record to check binary compatibility t20092.tmp ERROR    -1

解决：从上面看出，不识别 BEIDOU-3I，运行，grep 'BEIDOU-3I' *

antmod.dat:BEIDOU-3I           C38                 C220      2019-023A TYPE / SERIAL NO

antmod.dat:BEIDOU-3I           C39                 C221      2019-035A TYPE / SERIAL NO

antmod.dat:BEIDOU-3I           C40                 C224      2019-073A TYPE / SERIAL NO

grep: atl.list: No such file or directory

grep: atml.list: No such file or directory

grep: map.list: No such file or directory

grep: met.list: No such file or directory

sp3.log:WARNING:200427:1114:31.0 ARC/ertorb: Radiation scaling unknown for BEIDOU-3I       : use GPS Block IIF value

svnav.dat: C   220  38   0  BEIDOU-3I              2952000.     U      0.3000  2019 110  0  0  2100   1  0  0  # Max. yaw rate unknown

svnav.dat: C   221  39   0  BEIDOU-3I              2949000.     U      0.3000  2019 175  0  0  2100   1  0  0  # Max. yaw rate unknown

svnav.dat: C   224  40   0  BEIDOU-3I              2870000.     U      0.3000  2019 308  0  0  2100   1  0  0  # Max. yaw rate unknown

发现是北斗C38、C39、C40卫星天线文件svnav.data里面出现问题，索性将BEIDOU-3I改成BEIDOU-3M-CAST或者BEIDOU-3M-SECM，可以解决问题。（但是具体有什么差异，对结果精度有多大影响，还有待深入研究）

另外，北斗卫星轨道积分一直出现WARNING

WARNING:200427:2010:52.0 ARC/earthradtum: No Albedo model for SVN    5 Body BEIDOU-2I

WARNING:200427:2010:52.0 ARC/ertorb: Antenna thrust model not available for BEIDOU-2I

2）fixdrv dbdxt0.092

显示缺失g-文件或者t-文件，如果没有生成t-文件，那么在sh_sp3fit时候后加上参数-t，如果存在t-文件，还出现错误，那么估计是文件名没有统一，比如我的错误是dbdxt0.092中需要tgbm0.092，但是我真正生成的t文件名字是tgbm20993.sp3，所以只需要把tgbm20993.sp3改正过来就行了！

说明：以上出现的一些文件名字不对应，估计是分布解算的问题，用批处理sh_gamit不会出现这样的问题。

3）csh bbdxt0.bat

FATAL  :200427:1640:19.0 MODEL/open: Error opening IONEX file: ffile0 ERROR     2

这个错误显示，电离层模型文件IONEX出现错误。

IONEX（IONosphere Map EXchange）是最常用的空间电离层数据交换格式

在查看gamit10.71官方更新说明中，明确阐明sestbl.文件中默认使用IONEX文件

Ion model = GMAP                  ; Apply 2nd/3rd order ionospheric terms in model (default 'NONE')

Mag field = IGRF13                ; IGRF13/IGRF12/IGRF11/IGRF10/DIPOLE (default IGRF13)

利用sh_get_ion脚本下载Getting requested IONEX files

sh_get_ion -yr 2020 -doy 092

下载下来以后，IONEX文件名为：igsg0920.20i

而从上面错误信息IONEX file: ffile0，可以知道需要把igsg0920.20i改名为ffile0

然后重新解算（记得删除GAMIT.fatal）

电离层更为详细的信息，参考文献：

姜卫平2014《高阶电离层延迟对GPS坐标时间序列的影响分析》

4）csh bbdxt0.bat

解决上面问题以后，最后还是出现如下错误：

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite  2

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite  3

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite  4

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite  5

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite  6

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite  7

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite  8

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite  9

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 10

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 11

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 12

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 13

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 14

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 15

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 16

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 17

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 18

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 19

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 20

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 21

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 22

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 23

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 24

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 25

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 26

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 27

WARNING:200427:2031:29.0 SOLVE/dopt: No observation to satellite 28

STATUS :200427:2031:29.0 SOLVE/lsquar: Calculating new normal equation submatrices

STATUS :200427:2031:29.0 SOLVE/lsquar: Finding and removing dependent biases

STATUS :200427:2031:29.0 SOLVE/lsquar: Applying a priori 1000.0 cyc sigma on biases

STATUS :200427:2031:29.0 SOLVE/lsquar: Solving initial normal equations

STATUS :200427:2031:29.0 SOLVE/lsquar: Finished solving initial normal equations

STATUS :200427:2031:29.0 SOLVE/lsqerr: Constrained bias-free nrms =        NaN

STATUS :200427:2031:29.0 SOLVE/lcloos: Performing LC biases-free loose solution

FATAL  :200427:2031:29.0 SOLVE/lcloos: Inversion error in LCNORM(2)

这个原因，还没有找到，估计是星历的问题；尝试其他三个系统G、R、E都能正常处理获得结果；

目前提供北斗三号精密星历的只有武汉大学和GFZ两个机构，上面是GFZ处理的结果；

而换一个机构的星历，武汉大学的WUM尝试解算，出现如下错误：

/lib/rsp3hd: Error reading lines 8-12 of sp3c file ERROR 5010

这个错误应该是WUM精密星历的头文件中格式和GAMIT读取不一致造成；