Tag: GEOID12B

These columns have focused on procedures and routines for establishing GNSS-derived orthometric heights. There are many ways to analyze and investigate GNSS data and adjustment results. I have provided some basic concepts that I believe are important for users to understand.

The selection of constraints is a very important part of establishing accurate and consistent NAVD 88 GNSS-derived orthometric heights. All of the analysis and recommendations have been based on using the National Geodetic Survey‘s latest scientific geoid model.

I recommend first performing the analysis using the scientific geoid model because the hybrid geoid model has been warped to be consistent with the published NAVD 88 values. However, as mentioned in Part 7 (June 2016), in practice, GNSS-derived orthometric heights are incorporated into the NAVD 88 using the latest hybrid geoid model GEOID12B. This column will focus on the NGS “GPS on BMS (GPSBM)” dataset that was used to create the hybrid geoid model.

As mentioned in Part 3 (October 2015), the hybrid geoid model is designed to fit the published NAVD 88 leveling-derived orthometric heights. Saying that, the GPSBM dataset can be used to identify potential issues in the NAVD 88 published orthometric heights. GNSS users should be familiar with this dataset and how it can be used in their analysis. This column will provide tools and routines that can be used to identify potential issues in NAVD 88 heights and/or NAD83 (2011) published ellipsoid heights.

The National Geodetic Survey provides information on the bench marks occupied by GPS that were used to make GEOID12B.

The write up from the NGA website is given below. I have highlighted a few sentences that I’ll address in this column.

Table 1 is an excerpt of the excel spreadsheet for the GPSBM dataset and provides a sample of the contents. The headings of the columns are fairly self-explanatory. What’s important here is that the excel spreadsheet provides the name, latitude, longitude, NGS’ PID, the ellipsoid height and orthometric height of the stations used in making GEOID12B.

Table 1
Excerpt of the Excel spreadsheet for GPS on benchmarks (GPSBM) used to make GEOID12B.

The “GPS On Bench Marks (GPSBM) Used To Make GEOID12B” write up states that 1,721 stations were rejected and were not used in developing the hybrid geoid model. It also states that for the conterminous United States (CONUS), there were 24,782 stations with 911 rejected leaving 23,961. This column is going to focus on CONUS but the analysis can be performed everywhere.

As the write up states, stations were rejected for four different reasons:

• Code S – The State Advisor (now called Regional Geodetic Advisors) recommended it be dropped,
• Code h – The residual ellipsoid height errors from the NAD 83 (2011) project indicated that the point was too noisy,
• Code H – The orthometric height was suspect,
• Code N – The residual errors during geoid modeling were too high.

These rejected stations were not used to make the hybrid geoid model but since the hybrid geoid model is distorted to fit the NAVD 88, these rejected stations as well as stations nearby the rejected stations should be re-evaluated using the latest scientific geoid model, e.g. xGeoid16b.

So, what should the user do with the GPSBM table? I recommend that users perform the following steps when analyzing the stations in the GPSBM table.

• Step 1: Compare the modeled GEOID12B (N12B) value to the computed GPS/Leveling (h minus H) value using the following formula: Published N12B from the NGS data sheet minus (ellipsoid height from the GPSBM table minus orthometric height from the GPSBM table). We discussed this procedure a year ago in Part 3 (October 2015). It should be noted that the orthometric height in the GPSBM table may be different than the published NAVD 88 height on the NGS data sheet if the station has been readjusted since the GPSBM table was created.

• Step 2: Repeat the procedure in Step 1 using the latest NGS experimental geoid model, e.g. xGeoid16b. At this time, NGS only provides the experimental geoid models referenced to IGS08 so the user will have to use NGS’ xGeoid16 web tool to obtain the station’s IGS08 ellipsoid height and xGeoid16b value. The input to the tool is the station’s NAD 83 (2011) coordinates (latitude, Longitude, and ellipsoid height). [An example of using the xGeoid16 web tool is provided in the box titled “Example of Using NGS xGeoid16 Web Tool.”] As discussed in Part 3 (October 2015), the user will have to remove a bias and trend based on the differences in the region.

• The user could also transform xGeoid16b/IGS08 geoid values to xGeoid16b/NAD 83 (2011) geoid values using their own tools, and then remove a bias and trend based on the differences. Michael Dennis, a PhD candidate at Oregon State University, created an ArcGIS raster of the xGeoid16b model, where his model has been referenced to NAD 83 (Michael L. Dennis, RLS, PE, MS Civil Eng., Geodetic Analysis, LLC, 55 Creek Rock Road, Sedona, AZ 86351). He removed a trend using the GPS/Leveling data set as input; therefore, this raster file is a form of a hybrid geoid model distorted only to remove the tilt assumed to be in the NAVD 88. I will refer to this model as Geoid16B_NAD83 to avoid confusion with NGS’ xGeoid16b model.

• Step 3: Use the station’s data sheet to identify how the station’s orthometric height was determined; for example, was it rigorously adjusted into the NAVD 88 (published height attribute – Adjusted). We discussed the attributes of the NGS data sheet in Part 5 (February 2016). A summary of the attributes from the NGS data sheet DSDATA.TXT file is provided in the box titled “Extracted from NGS’ DSDATA.TXT.” I have highlighted the most common attributes of the stations involved in making GEOID12B.

• Step 4: Use the station’s NGS data sheet to determine the adjustment date of the station’s published NAVD 88 orthometric height. We discussed this in Part 7 (June 2016). As mentioned in Part 7, if the station has a different adjustment date than other stations nearby, there could be inconsistencies due to adjustment distribution corrections and/or movement.

Step 1 was demonstrated in Part 3 (October 2015) so we don’t need to describe the process in this column. Comparing published GEOID12B values with computed values is the first step; the difference is an indication of how well the data fit the model and can be useful for identifying large outliers. It can be helpful in prioritizing where additional observation should be obtained when there are limited resources. Provided below is an example of where to obtain the information for comparing the modeled GEOID12B (N12B) value to the computed GPS/Leveling (h minus H) value using the following formula: Published N12B from the NGS data sheet minus (ellipsoid height from the GPSBM table minus orthometric height from the GPSBM table). The user can obtain the GEOID12B value from the NGS data sheet [see box titled “Excerpt from NGS Data Sheet For Station L 275 (HW2088)”]; for this example, the GEOID12B value for station L 275 is -30.813 m. Table 2 is an excerpt from the GPSBM file that contains the ellipsoid height (599.253 m) and the orthometric height (630.016 m) for station L 275. It should be noted that the ellipsoid and orthometric heights in the GPSBM table are given in millimeters. The first row of table 3 provides the results of the computation: [-30814 mm – (599253 mm – 630016m m) = 51 mm], or 5.1 cm.

Table 2
Excerpt of the Excel spreadsheet for GPS on benchmarks (GPSBM) used to make GEOID12B – Stations on plots in this column.

Table 3 contains the comparisons between modeled geoid values and their computed geoid values for five station pairs that have large relative differences. Looking at table 3 one can see that there are several large relative differences between the published GEOID12B model and computed geoid model (see column titled “N12B minus (h-H)” in table 3). This doesn’t mean that the model is incorrect, it only means that there were large relative differences that the model had to account for. As previously mentioned, GEOID12B was created to be consistent with the NAVD 88.

Since the experimental geoid model xGeoid16b_NAD is not distorted to conform to the NAVD 88 everywhere, it should provide better information for identifying outliers and determining which stations appear to be inconsistent with its neighbors.

Table 3
Table of selected stations involving large relative differences depicted in plots in this column.
(Results are provided for GEOID12B and Geoid16B_NAD Models*)
*Michael Dennis, a Ph.D. candidate at Oregon State University, created the xGEOID16B ArcGIS raster, where the model has been referenced to NAD 83 with a trend and bias added to account for the apparent tilt in the NAVD 88. This model is denoted as Geoid16B_NAD (N16b) in this column.

Figure 1 is a plot of all of the GPSBM residuals using the Geoid16B_NAD83 model. This plot indicates that there are a lot of large residuals. First, let’s define what I’m calling residuals. The residuals on my plots are the differences between the modeled geoid height value and the computed geoid height value using the ellipsoid height (h) and orthometric height (H) from the GPSBM data set; that is, residual = modeled gravity value – (h minus H). The largest negative residual is -37.3 cm and the largest positive residual is 33.8 cm.

Figure 2 is a plot of the positive GPS on BMS residuals using Geoid16b_NAD geoid model. There are 5957 residuals greater than 5 cm (not including the stations rejected by the NGS geoid team). As you can see, it appears that most of the positive residuals are on the eastern half of the United States.

Figure 3 is a plot of the negative GPS on BMS residuals using Geoid16b_NAD geoid model. There are 4113 residuals less than -5 cm (not including the stations rejected by the NGS geoid team). As you can see from the plot, the negative residuals appear to be more evenly distributed across the United States than the positive residuals. It does, however, appear that there are more negative residuals greater than -5 cm along the Gulf Coast, Atlantic Coast, and the Great Lakes than there are positive residuals greater than 5 cm. In addition, there appears to be a lot of negative residuals in the northeastern United States.

Figure 4 is a plot of the GPS on BMS residuals using the Geoid16b_NAD geoid model in the North Carolina and South Carolina border region. What’s interesting about this plot is that South Carolina doesn’t seem to have many negative residuals where North Carolina has both negative and positive residuals. We will look at this in more detail later in this column.

Figure 5 is a plot of the GPS on BMS residuals using Geoid16b_NAD model in the Washington and Oregon Region. This graphic shows some large grouping of negative and positive residuals, especially along the Pacific Coast in Northwestern Washington State.

Now, let’s look at some large relative differences in residuals between stations that are spatially close together. Figure 6 is a plot of large relative differences between groups of GPS on BMS residuals (using Geoid16b_NAD model) at the North Carolina/South Carolina border. In figure 6, two stations (FA1337 and FA1560) are about 20 km apart and the difference in residuals is -18.6 cm (-12.4 cm minus 6.2 cm). This is a large difference for only 20 km. What is even more significant is that the group of stations near FA1337 are all negative residuals (around -10 cm) and the group of stations near FA1560 are all positive residuals (around 6 cm), this could be an indication of a large distribution correction due to the NAVD 88 design. We discussed the distribution correction in Part 7 (June 2016). These stations definitely needs to be investigated.

The next step in my process is to look at the NGS data sheets for these stations to determine how the stations were adjusted.

The data sheet for station FA1337 states that the NAVD 88 attribute code is “GPS OBS.” [See box titled “Excerpt from NGS Data Sheet for PID FA1337.”] The data sheet for FA1560 states that the NAVD 88 attribute code is “Adjusted.” The orthometric height on the GPSBM file is different than the current published NAVD 88 orthometric height for station FA1337 (See table 3). This station’s leveling-derived orthometric height was superseded by a GNSS-derived orthometric height. Saying that, the GPSBM file only uses leveling-derived orthometric heights; therefore, stations that have been superseded by GNSS surveys are still included in the GPSBM file but their original published leveling-derived height is used for the analysis. Table 3 provides the orthometric height for FA1337 that was used in making GEOID12B. As previously mentioned, stations may be rejected by the geoid team based on the criteria outlined in the beginning of this column. Saying that, neither of the two stations were rejected by the NGS geoid team. This implies that the stations were consistent with their neighbors as far as the geoid model was concerned. Figure 6 confirms that all the stations around FA1337 and FA1560 are consistent with each other based on the Geoid16b_NAD geoid model. The fact that the two groups differ by 18 6 cm needs to be investigated.

Figure 7 is a plot of the GPS on BMS residuals using Geoid16b_NAD that depicts a large difference between two stations only 20 km apart near the Maryland/West Virginia border. I will use this station pair to demonstrate the next step in my process.

The NAVD 88 attribute on the NGS data sheet states that both of these stations are coded as “Adjusted” but station JW0639 adjustment date is April 1995 (see box titled “excerpt from NGS Data Sheet for PID JW0639”) and JW1296 adjustment date was in June 1991 (the General Adjustment of NAVD 88). These large relative differences could be due to inconsistencies between adjusted heights due to the adjustment distribution corrections and/or constraints imposed in the April 1995 adjustment. Bench marks near the stations should be observed to determine if the same large relative difference exists, and the 1995 NAVD 88 adjustment project report should be reviewed to determine if a large distribution correction was applied.

Figure 8 is a plot of GPS on BMS residuals using Geoid16b_NAD that depicts a large relative difference between stations 15 km apart in Randolph County, West Virginia. This plot involves station HW3677 which has a published NAVD 88 attribute of “Leveling.” (See box titled “Excerpt from NGS Data Sheet for PID HW3677.”) The excerpt from the data sheet has the following statement: “The orthometric height was determined by differential leveling. The vertical network tie was performed by a horz. field party for horz. obs reductions. Reset procedures were used to establish the elevation.”

It would be useful if stations near this station were observed by GNSS surveys to determine what is occurring in this region.

Figure 9 is a GPS on BMS residual plot of large relative stations about 30 km apart in Wasco County, Oregon. This plot has two stations with large differences and both stations have the NAVD 88 attribute of “Adjusted.” Their NGS data sheet states that they were both established in the general adjustment of NAVD 88 in June 1991. In this particular case, the leveling in this region is very old. As described in Part 7 (June 2016), you can retrieve all project identifiers for those projects with observations to or from a station using the station’s PID. The output from the NGS Data Sheet Mark Source Routine for PID RC1228 is shown in the box titled “Output from NGS Data Sheet Mark Source Routine.”

Figure 9 is a GPS on BMS residual plot of large relative stations about 30 km apart in Wasco County, Oregon. This plot has two stations with large differences and both stations have the NAVD 88 attribute of “Adjusted.” Their NGS data sheet states that they were both established in the general adjustment of NAVD 88 in June 1991. In this particular case, the leveling in this region is very old. As described in Part 7 (June 2016), you can retrieve all project identifiers for those projects with observations to or from a station using the station’s PID. The output from the NGS Data Sheet Mark Source Routine for PID RC1228 is shown in the box titled “Output from NGS Data Sheet Mark Source Routine.”

Figure 10 is a plot of GPS on BMS residuals using Geoid16b_NAD depicting large relative differences between stations along the Oregon/Washington State border. It is the near Puget Island along the Columbia River. Station SC0330 and SC1086 are only 7 km apart and the relative difference is -20 cm (-11.4 cm minus 8.6 cm). This could be an issue with the NAVD 88 network design because there doesn’t appear to be many river crossing along the river between border stations. The fact that the residuals on the Washington State side are negative and the Oregon State side are positive is an indication that the stations need to be investigated.

The last figure, figure 11, is a plot of the GPS on BMS residuals using Geoid16b_NAD model that depicts large negative residuals north of the border between Oregon and Washington and positive (or small negative) residuals south of the border. This plot shows that the northern side of the river has large negative residuals all the way to the Pacific Coast. Once again, this is an indication that this portion of the NAVD 88 network should be investigated.

This column has focused on analyzing NGS’ GPS on BM data set that is used to make NGS’ hybrid geoid models. It provided procedures that users could employ when analyzing the differences between the modeled geoid values and the computed geoid values using GPS/Leveling data. This GPSBM data set or one similar will be used to make the next hybrid geoid model, as well as provide input to the transformation model between NAVD 88 and the new 2022 Vertical Reference System. All geospatial users should help develop this GPS on BMS data set to help improve the National Spatial Reference System and future hybrid geoid models. This column provided several examples of large relative differences in residuals between neighboring stations. Each example represents stations that should investigated based on different reasons, such as a weak NAVD 88 leveling network design in the region, the station’s published height attribute code implies that the station was not rigorously adjusted into the NAVD 88, and station pairs have different adjustment dates indicating a possible adjustment distribution correction issue or movement.

NGS has a program called “GPS on Bench Mark” to support users that occupy bench marks with GNSS equipment. This web site contains a lot of good information and provides the users with methods to recover, observe, and report information about stations in NGS’ database. The write up from the webpage is given below. I have highlighted a few sentences that the reader may find useful.

In addition to participating in the NGS’ GPS on Bench Mark program, all geospatial users should participate in NGS’ 2017 geospatial summit, which will be held in April in Silver Spring, Maryland.

This summit is an opportunity for all users of the National Spatial Reference System (NSRS) to obtain a better understanding of NGS’ plans to modernize the NSRS. Users will be able to provide feedback directly to NGS leadership. My next column will address NGS plans to replace the North American Vertical Datum of 1988 in 2022.