DISCUSSION OF THE EFFECTS OF SUBSIDENCE ON FLOODING AUGUST 8, 1995 by Michael D. Talbott, P.E. Engineering Division Manager HARRIS COUNTY FLOOD CONTROL DISTRICT, TEXAS OVERVIEW Three primary issues are presented in this discussion: 1) 2) 3) The effective Flood Insurance Study for Harris County, Texas (September 30, 1992) contains an appropriate discussion of policy on subsidence, and the discussion was purposely placed in the Differential subsidence in inland areas of Harris County is not extreme enough to warrant concerns over changes in the depth or extent of flooding; Due to relatively uniform subsidence, it is not necessary to use current and accurate ground elevations in inland areas since the Reference Marks (benchmarks) subside relative to the flood elevations. SUMMARY OF KEY POINTS The following points are made regarding the three primary issues presented. above. The points are restated in bold characters throughout the text. The Harris County Flood Control District believes the subsidence policy in the effective FIS is appropriate, and that it was purposely placed in the FIS. The policy as submitted to, reviewed by, and adopted by Harris County is technically sound, logical, and reflects the results of ,years of discussion, technical analyses, and practical application to numerous projects. The key finding the Subsidence/Flooding Study for inland riverine flooding conditions was that differential subsidence can affect the depth of flooding either positively or negatively depending (Hi the resulting effect (Hi the overall stream gradient. The groundwater control plan is very effective in controlling subsidence. All of the stable benchmarks (extensometers) have experienced a significant leveling off of the rate of subsidence. The extreme scenarios analyzed in the Subsidence/Flooding Study showed that the net effect of differential subsidence on the depth of flooding was only a fraction of the subsidence amount (either positively or negatively). When considering policy decisions regarding the effects of subsidence in Harris County, it is extremely important to be aware of the regional nature of the phenomenon. It is also vitally important to consider the drainage network characteristics relative to the pattern of subsidence. The lower aquifer pressures and wide-spread groundwater withdrawal cause general subsidence of a: large area (several thousand square miles). Drainage basins le the region are typically small relative to regional subsidence patterns, whereby the entire drainage basin is affected, not just isolated pockets. Given the extremes of differential subsidence that are required to produce a marked change in the depth and extent of flooding and the success of the groundwater control plan, there is no cause for concern at this time (or the foreseeable future) to warrant the drastic change in policy interpretation being proposed by FEMA. Due to the regional nature of the subsidence phenomenon and the extensive coverage of Reference Marks or benchmarks), benchmarks within reasonable proximity of the stream will maintain their relative elevation differential: The benchmarks subside along with the BFE. There is no need to use current and accurate elevations in inland areas if the relative elevations are consistent. The requirement to use current subsided ground elevations with unsubsided is not supportable technically given everything known about the subsidence/flooding relationship in Harris County. It is not technically feasible to expect isolated public works projects, flood control projects, or land. development activities to update isolated portions of the FIS models. It simply will not work because the regional nature of subsidence affects the entire stream system. The desire to update the models for subsidence and to convert to the NAVD can both be addressed when the entire county FIS needs to be updated. The useful life of the FIS is affected by a number issues, such as subsidence, changes in the HEC program, calibration to more observed flood data, existing errors 5J1 data cur calculations, natural changes 5J1 channel conveyance, possible conversion to the metric system, etc. The update should be considered for the future whereby the scope and financing of the effort can be planned for by the appropriate agencies. 0 In the meantime, policy and procedures should be documented that will allow the continued processing of the many projects that are generated from this region. The policy and procedures must rely on a reasonable application of the knowledge gained from the relationships established in the Subsidence/Flooding Study, watershed-specific conditions, and the past eight years of successful practical applications. 0 In. all cases, a registered. professional engineer should be required to evaluate the subsidence conditions of the area and certify/seal their findings and recommendations. ISSUE 1: THE EFFECTIVE FLOOD INSURANCE STUDY FOR HARRIS COUNTY, TEXAS (SEPTEMBER 30, 1992) CONTAINS AN APPROPRIATE DISCUSSION OF POLICY ON SUBSIDENCE, AND THE DISCUSSION WAS PURPOSELY PLACED IN THE FIS Recent letters from the Federal Emergency Management Agency (FEMA) state that the currently effective Flood Insurance Study (FIS) for Harris County, Texas (dated September 30, 1992) contains an incorrect discussion of the policy on subsidence and that the policy was mistakenly placed in the FIS. The Harris County Flood Control District believes the subsidence policy in the effective FIS is appropriate, and that it was purposely placed in the FIS. The policy as submitted to, reviewed by, and adopted by Harris County is technically sound, logical, and reflects the results of years of discussion, technical analyses, and practical application to numerous projects. The following' summary' of events shows ?the logical evolution to the currently effective FIS wording on subsidence: 0 The preliminary FIS obtained by in March 1983 was prepared for the Corps of Engineers? study and contained a (3 page, 2 exhibit) discussion of how to View subsidence in coastal and inland areas. In essence, the discussion said to compare the Base Flood Elevations (BFE's) based, on 1973 releveling of the .National Geodetic 'Vertical Datum (NGVD) from the FIS with 1973 adjusted ground elevations in inland areas, and to the most current ground elevation in coastal areas. For inland areas, this approach presumes that the water surface subsides at the same rate as the land, and the depth of flooding and flood plain limits remain constant. In areas subject to tidal flooding, flood levels would remain constant as the land subsides which would increase the depth of flooding and the limits of the flood plain. 0 The preliminary FIS narrative received in July 1984 condensed the discussion of subsidence to three short paragraphs. The policy said to use of current ground elevations with the effective in ?the for' all conditions, coastal and inland. 0 Many discussions followed between FEMA, and Dewberry Davis (FEMA's technical evaluation contractor). A meeting was held with FEMA on October 3, 1984 where recommended that FEMA differentiate between coastal and inland flooding in a: manner similar to that proposed in the March 1983 preliminary FIS, and that appropriate subsidence considerations would be established (n1 a project?by-project basis by the local Flood Plain Administrator and/or FEMA. 0 provided review comments on the preliminary FIS for the City of Houston and Unincorporated Harris County to FEMA by letter' dated, December 13, 1984. The letter brings up the unresolved issue of subsidence as one of major concerns. 0 Further discussions lead to the initiation of negotiations between four local agencies to fund and administer a study of the relationship between subsidence and flooding. A. joint venture of three local engineering companies was selected to conduct the study, with funding coming from the Fort Bend County Drainage District, Harris County Flood Control District, Harris-Galveston Coastal Subsidence District, and City of Houston. 0 FEMA responded to Harris County Judge Jon lindsay by letter dated March 25, 1985. The letter commented on subsidence: The: question of land subsidence and its effects on flood levels has been discussed in conversations and correspondence, between. representatives of the Harris County Engineering Department, and FEMA over the past year. It is a complex question not solvable without further technical analyses. However, we understand that such analyses have been initiated by Harris County in cooperation with other local government agencies. FEMA applauds this effort and is committed to working with Harris County in developing procedures for implementation of the analyses when completed. Until such time, FEMA will work with Harris County in establishing interim guidelines and procedures for utilizing the FIS for Harris County in the face of ongoing ground subsidence. The FIS was released in 1985 with the short vershmu of the subsidence policy whereby subsidence is applied to the ground elevations but not the 0 As a condition for continuing in the National Flood Insurance Program, Harris County was required to update its flood plain regulations to reflect the detailed study (FIS) information. FEMA reviewed and Harris County adopted its flood plain management regulations which define "elevation" as follows: "Elevation" means height above mean sea level. The 1973 National Geodetic Survey re~leveling of the vertical control system (benchmarks) referenced to 1929 mean sea level datum shall be used except in coastal areas where subsidence has occurred. The most recent releveling of the vertical control system shall be used in the coastal areas. Any future studies changing the FIRM which is referenced to a later re?leveling of the vertical control system shall be used whenever a revised FIRM becomes effective. 0 The subsidence/flooding study was completed and entitled Study Of The Relationship Between Subsidence And Flooding." Copies were provided to FEMA. The study supports the concept that there is a difference between inland and coastal flooding, and establishes the relationship that significant differential subsidence must occur in inland areas before the depth or extent of flooding are affected. 0 More discussions were held between the various agencies and engineers, and informal operating guidelines and procedures were developed and implemented. In practice, the engineers (FEMA, and Dewberry Davis) considered the location (coastal or inland) and the relative uniformity of subsidence in a watershed to decide if subsidence was an issue. The course of action on many projects was decided using the established procedures. The conversion of the community-based FIS mapping to the county-wide format was ongoing during the time-frame of these discussions and decisions. 0 The preliminary FIS for Harris County was released in 1989 in the county-wide format with the expanded discussion of inland and coastal considerations restored. The local community reviewed the document and concurred with the presentation as being technically supportable (based on the Subsidence/Flooding Study), practical (based on the years of application on many projects), and logical. The policy was essentially the same discussion (3 pages, 2 exhibits) that was seen in the preliminary FIS in 1983; 0 The FIS went through the statutorily-required due process of public notice and appeals. 0 The 1990 F15 was issued and contained what the community considered an appropriate subsidence policy given the extensive discussions, engineering analyses, and six years of practical application. 0 The 1992 F18 and 1994 preliminary FIS were issued with the same appropriate subsidence policy. In summary, the progression of discussions, engineering analyses, practical application to many projects over many years, and public due process lead to the current wording in the effective FIS. It was not a mistake that the policy appeared in the 1989 preliminary FIS: that was the first republication of the FIS for Harris County after the detailed engineering study of subsidence and flooding was completed. As stated in the March 25, 1985 letter from FEMA: applauds this (study) effort and is committed to working with Harris County in developing procedures for implementation of the analyses when completed . This is exactly what was done. ISSUE 2: DIFFERENTIAL SUBSIDENCE IN INLAND AREAS OF HARRIS COUNTY IS NOT EXTREME ENOUGH TO WARRANT CONCERNS OVER CHANGES IN THE DEPTH OR EXTENT OF FLOODING Background Subsidence has been occurring in the Houston metropolitan area for about 100 years. The primary cause of subsidence in this region is the withdrawal of groundwater. Subsidence in some coastal areas has lowered ground elevations relative to sea level where the effect on flooding is obvious: more permanently inundated land and more land subject to flooding (areal and depth) from tidal surge associated with tropical storms. Recognizing the impact of subsidence on coastal flooding, the State legislature in 1975 created the Harris-Galveston Coastal Subsidence District (H-GCSD) to plan and regulate groundwater withdrawal. The H-GCSD's primary function is to develop and implement a plan to regulate groundwater withdrawal to control subsidence. Coastal subsidence has been brought under control through the successful implementation of the groundwater control plan. As continued conversion to surface water is implemented under the plan, the rate of inland subsidence is diminishing as well. Historically, the majority of regional subsidence was concentrated near the early development and industrial areas along the Houston Ship Channel. The Ship Channel serves as the primary conduit for flood waters for much of the central Houston/Harris County area. The ihistoric subsidence ;patterns generally increased the gradient of tributaries to the Ship Channel, which was (correctly) believed to actually benefit inland drainage and flooding. Although control of coastal subsidence has been achieved, increasing westward urban growth and related water supply needs have resulted in continued inland subsidence. Inland subsidence toward the west had the potential to adversely affect the stream gradients, as well as affect the flood protection capabilities of the regional Addicks and Barker flood control reservoirs which protect Buffalo Bayou and downtown Houston. The effect of inland subsidence on flooding was not fully defined. The need for more definitive information became evident as the local entities moved forward in planning for water supply, drainage and flood. control, and groundwater regulation. To respond to the need for better information, a study was undertaken by the local entities primarily responsible for water supply, control of subsidence, and flood control in the Houston metropolitan area: Fort Bend County Drainage District; Harris County Flood Control District; Harris-Galveston Coastal Subsidence District; and City of Houston. The study is dated December 1986 and entitled Study Of The Relationship Between Subsidence and Flooding" "Subsidence/Flooding Study") (Ref. 1). The scope of the study included three major components of the drainage and flood control systems in the greater Houston area: riverine systems, localized drainage systems, and the Addicks and Barker flood control reservoir system. The riverine flooding analysis portion of the study is an evaluation of flooding that may result from potential subsidence along main drainage channels with the objective of determining if a relationship exists between {gradient change caused by subsidence and storm flows and/or flood plain area. The localized drainage analysis is an evaluation of the impacts of regional subsidence and well field placement on localized drainage systems and the resultant effects on minor drainage channels, storm. sewer systems, and street pending. The reservoir capacity analysis addresses the effects that subsidence-caused gradient changes have on the maximum flood storage capacities and loo-year pool levels of the Addicks and Barker flood control reservoirs. The key finding of the Subsidence/Flooding study .for inland riverine flooding conditions was that differential subsidence can affect the depth of flooding either positively or negatively depending on the resulting effect on the overall stream gradient. It should also be pointed out that the scenarios analyzed in the study were generally for extreme differential subsidence conditions. The extreme scenarios are based on the physical conditions of the aquifer system, but the extreme scenarios do not recognize the current requirements for decreases in groundwater pumping under the H-GCSD plan which have and will continue to decrease the realized subsidence. The actual observed subsidence is substantially less than the scenarios analyzed. Extreme scenarios were used 113 envelope physically possible conditions to establish the relationship of subsidence and flooding. Analyzing conditions that would essentially abandon the H-GCSD groundwater control plan was important also in order to consider if groundwater should be controlled in inland areas at all since there is a cost to convert to surface water. Success of Groundwater Control Plan The plan for groundwater control has been very effective in controlling subsidenCe. While the overall water supply requirements for the region have increaSed from 738 million gallons per day (MGD) in 1976 to 877 MGD in 1994, the percentage of the supply coming from groundwater has decreased from 456 MGD to 319 MGD (Ref. 2). The result has been an ever decreasing rate of subsidence at all of the permanent stable benchmarks maintained by the H-GCSD. Figure 1 is a chart of cumulative subsidence at the 12 permanent borehole extensometers (stable benchmarks) maintained by the for the period of 1973 to 1994 (Ref. 2). The location of the benchmarks are shown on Figure 2 (Ref. 2). Several points are made based on Figure 1: 0 The H-GCSD's groundwater control plan is very effective in controlling subsidence. All of the stable benchmarks (extensometers) have experienced a significant leveling off of the rate of subsidence. The most westerly benchmark (Addicks) has seen less of a leveling off, but also has a decreasing rate of subsidence. Subsidence rates on the graphs can be characterized as steep, milder, and flat, which generally correspond to ?three time periods: 1973?1978; 1978-1987; and 1987-1994. These time periods reflect the implementation and success of the H-GCSD's groundwater control plan. The plan was first implemented in 1976 (and has been updated since). Characteristic Period Subsidence Rate 1973-1978 Steep 1978-1987 Milder 1987-1994 Flat 0 Several letters received recently from FEMA have raised concerns about subsidence which may have occurred since the last general region-wide releveling effort (1987). The concern is illustrated in the letters by extrapolating the subsidence between 1978 and 1987 to what might have happened from 1987 to 1995. FEMA's illustration. is inappropriate considering' the significant stabilization of the benchmarks as illustrated by the graphs on Figure 1. 0 Eleven out of twelve locations have had virtually no subsidence in the last 5 years (some have actually showed some minor recovery). i General SubsidencelFlooding Relationship Coastal subsidence is a cdefinite problem ?where depth, of flooding increases nearly the same as the amount of subsidence (S) (given that the overall subsidence pattern is not significant enough to change the coastal surge analysis). The extent of flooding also increases. 1. INCREASED REACH OF I INUNDAHON (8) Coastal Flooding Schematic Inland subsidence is a different phenomenon. a. Where subsidence (S) is uniform (SA SB) such that the slope of the channel system is unchanged, the extent and depth of flooding will not change. Flows also do not change. 2. Inland Uniform Subsidence Schematic has?: Where subsidence (S) is not uniform and the slope of the channel system is affected, the depth of flooding may be affected. This case results in differential subsidence (A) along' the channel system. The term differential subsidence is important to this discussion. Given the regional nature of the subsidence phenomenon in Harris County and the relatively small size of the drainage basins, the entire watershed is affected by subsidence (more on this later). As discussed above, if the entire basin subsides the same amount, there is no change in the flow, depth or extent of flooding. To introduce an impact, differential subsidence must occur such that the gradient of the stream is affected. The measurement of the ?differential is :made within the context of the entire affected reach of the stream (A SB - SA). As the Subsidence/Flooding Study shows, this differential subsidence must be significant in order to affect the depth and extent of flooding (more on this later, too). - Differential subsidence (A) which results in steepening of the stream gradient yields a shallower depth of flooding Flows increase but by an amount less than the increase in conveyance capacity. Inland Steepening Differential Subsidence Schematic 10 "svxnew - Differential subsidence (A) which results in a flattening of the stream gradient yields an increases in depth of flooding Flows decrease but at a rate slower than the loss in conveyance capacity. Inland Flattening Differential Subsidence Schematic Differential subsidence (A) which results in a combined steepening of the upper reach with flattening in the lower reach results in shallower depth of flooding in the steeper section (du) and increased depth of flooding in the flatter reach Flows increase in the steeper reach and decrease in the flattened reach. dd2>dd1 AU: Inland Combined Differential Subsidence Schematic 11 Stream Gradient Specific Findings of the Subsidence/Flooding Study The extreme scenarios analyzed in the Subsidence/Flooding Study showed that the net effect of differential subsidence on the depth of flooding was only a fraction of the subsidence amount (either positively or negatively). For cases where resultant gradients were steeper, the average net change in the depth of flooding was about -0.07 foot per foot of differential subsidence. For this type of effect to result in a net decrease in the depth of flooding of 0.5 foot, a total differential subsidence of about 7.1 feet would have to occur. For cases where flatter gradients were realized (either from an overall flattening or a combined steepening and flattening), the net change in the depth of flooding was on average 0.1 foot per foot of differential subsidence. The worst case produced 0.3 foot. per foot of ?differential subsidence. For this type of effect to result in a net increase in the depth of flooding of 0.5 foot, a total differential subsidence of about 1.7 to 5.0 feet would have to occur. Exhibit II-lz of the Subsidence/Flooding Study (Ref. 1) shows this relationship graphically and is reproduced for this discussion as Figure 3. Change In Depth/Foot of Differential Subsidence Differential Subsidence to Net 0.5 Foot Change Steepened ?0.07 (average) 7.1 Flattened +0.1 (average) 5.0 +0.3 (most) 1.7 While 1.7 feet of differential subsidence does not seem extreme, it is important to note the details of the three (out of 48) cases that produced the highest adverse effect (up to 0.3 foot increase in depth of flooding per foot of differential subsidence). 0 The ?three cases analyzed. were on Brays Bayou (Case. 1) and Buffalo Bayou (Cases 1 and 2). The combination scenarios (Cases 1 and 2) of steepening in the upper part of the drainage area and flattening in the lower part resulted in differential subsidence on the order of 2.9 to 7.4 feet flattening and 3.5 to 8.1 feet steepening. The maximum subsidence at a point which produced the differentials in these scenarios was 9.8 feet. 0 The scenarios were extreme conditions of what is physically possible but won't occur due to the existence of the H-GCSD and their authority to regulate groundwater under their ongoing plan of conservation and conversion to surface water. These are not real conditions, only extremes to establish relationships. 12 0 Case 1, which produced two of the extreme results, is the scenario where the center of the subsidence condition was placed at a point where 80% of the drainage area is upstream and 20% is of that point. This scenario produced higher flows and lower depths of flooding upstream. of the center (a significant reach of the channel) and lower flows with higher depths (a much shorter reach), with the largest impact occurring just of the center. The other case (2) placed the center of the subsidence condition at a point where 50% of the drainage area was upstream and 50% was 0 Of greater significance is the degree of change in stream gradient over the available stream gradient. Both Brays and Buffalo bayous are relatively flat stream systems with overall gradients of about 2.1 to 3.2 feet per mile. The differential subsidence modeled to pmoduce these "worst results" produced flattening of the gradients by 10% to 20%, and steepening of the upstream gradients by 13% to 18%. (Again, these extremes do not recognize the ongoing successful implementation of the H-GCSD groundwater control plan). It is also important to examine the data presented on Exhibit II-11 (ref. 1) which relates the change in depth of flooding to the percent change in stream gradient. Exhibit is reproduced for this discussion as Figure 4. In order to create an average 0.5 foot change in the depth of flooding, the percent change in slope of the stream is extremely high. For cases where resultant gradients were steeper, it took at least a 20% increase in the slope to decrease the depth of flooding' by 0.5 foot. Under' the flattening" of the. gradient scenario, the most sensitive stream (Brays Bayou) required a 9% decrease of the gradient to increase the depth of flooding by 0.5 foot. The other streams required an even more drastic change in slope on the order of 30% to 50%. Again, the two most extreme results were on Brays Bayou and Buffalo Bayou under the conditions discussed above. Percent Change In Gradient Stream Gradient to Net 0.5 Foot Change in Depth Steepened Flattened (Most) 9% (Average) Regional Nature of Subsidence When considering policy decisions regarding the effects of subsidence in Harris County, it is extremely important to be aware of the regional nature of the phenomenon. It is also vitally important to consider the drainage network characteristics relative to the pattern of subsidence. 13 Beneath the Houston-Galveston area are a series of sand and clay beds which provide water (aquifers). Subsidence in the Harris County' region is created by excessive pumping? of groundwater which lowers the pressure in ?the aquifers. As the (pressure decreases, the clay? particles begin to collapse resulting in compaction (subsidence). The lower aquifer pressures and wide-spread groundwater withdrawal cause general subsidence of a large area (several thousand square miles). Drainage basins in the region are typically small relative to regional subsidence patterns, whereby the entire drainage basin is affected, not just isolated pockets. Primary drainage basins are on the order of 50 to 300 square miles. Figure 5 shows the results of the latest regional releveling survey as subsidence patterns for the region (1978*1986) (Ref. 3). The 1986 information is the latest regional releveling that allows the regional subsidence pattern to be plotted. A map showing the patterns for the period 1973-1987 is not readily available, and region-wide data for the period 1987-1995 is not available, yet. Additional subsidence has occurred since 1987; however, the overall patterns of controlled coastal subsidence and relatively slight inland subsidence have continued given the implementation of the plan for groundwater control (see Figure 1). The primary drainage basins are also shown on Figure 5 by vectors representing the relative size of the basin and direction of flow. As shown by the vectors, the entire basins are influenced by the regional nature of subsidence. Many of the basins have an overall (beneficial) steepening. Note that the differential subsidence (steepening, flattening, or combined) within a drainage basin is very slight and not sufficient to affect the depth or extent of flooding as determined by the relationships in the Subsidence/Flooding Study. Again, this data only shows results through 1987, but the general trends are consistent under the groundwater control plan (see Figure 1). Also shown on Figure 5 is the extent of the county that may be influenced by coastal flooding and therefore sensitive to subsidence effects. Special consideration must be given to the effects of existing and projected subsidence in these areas, but not the remainder of the county. The effectiveness of the groundwater control plan has brought coastal subsidence under control whereby the concern for subsidence-induced flooding is less pronounced. Specific Examples for Brays and White Oak Bayous Most of the FIS is based on data collected on the 1973 adjustment of NGVD. Differential subsidence that has occurred since the field data was collected has not been anywhere near the extreme scenarios devised to establish the relationship between subsidence and flooding. This can be attributed to the success 14 of the plan for controlling groundwater pumping through conservation and conversion to surface water. As examples of the slight differential subsidence, recent update/correction studies were performed on Brays Bayou and White Oak Bayou. The maximum subsidence within each basin between 1973 and 1987 (the date of the last general releveling of benchmarks) has been about 2.6 feet. However, due to the regional nature of the subsidence phenomenon, the entire drainage basins have also subsided. The maximum differential subsidence rates experienced are on the order of 1.0 to 1.6 feet. The percent change in stream gradient is also very slight. Example Changes in Gradient: 1973-1987 Average Existing Change in Percent Stream Gradient (ftimi) Gradient (ftlmi) Change Brays (steeper) 3.15 0.11 3.4% (Flatter) 3.15 -0.08 White Oak (Steeper) 5.28 0.12 2.2% (Flatter) 5.28 -0.06 *Note that the gradient change is about one inch per mile. Although the differential subsidence for Brays Bayou is considered very slight, ground elevations were updated from 1973 adjustment of NGVD to the 1987 adjustment. This amount of differential subsidence was not considered severe since the Subsidence/Flooding Study had done extensive modeling of Brays Bayou, including conditions that reflected implementation of the plan and the resultant subsidence between 1973 and 2020. This case was designated Case 3a in the study and produced differential subsidence on the order of 4.1 feet (flattening) and 1.0 foot (steepening). The results were barely perceptible changes in depth or extent of flooding in a very minimal reach of the channel. The flood plain results of the Brays Bayou 1973- 2020 case (3a) were mapped on Exhibits 11?18 and 19, and the plotted stream profile was presented on Exhibits II-20 to 25 (Ref. 1). White Oak Bayou elevations were not changed from the 1973 adjustment of NGVD because of the very small differentials and a steeper available gradient. (The change in gradient was on the order of one inch per mile over an existing gradient of over 5 feet per mile). This is also a convenient place to discuss an analysis of Brays Bayou that was performed by Baker Engineers and discussed at our July 13, 1995 meeting in Corpus Christi, Texas. Baker Engineers gave a presentation at the Corpus Christi seminar on their 15 analysis of the effects of subsidence on flooding using Brays Bayou as ?n1 example. considers the analysis incomplete, and therefore inconclusive. Of primary concern is the fact that discharges were not recalculated to reflect the relationship that flows decrease when the stream gradient is flattened. Also, the analysis only looked at the portion of the stream that was flattened, and then only emphasized the point on the stream where the results were the most extreme. Baker Engineers also made a statement to the effect that "the recent Brays Bayou study produced similar results to their analysis." The study included many changes in hydrologic and hydraulic parameters, not just subsidence. Due to the many changes in the models that were done in the analysis, it is impossible to isolate the effects of subsidence alone. The Baker Engineers' incomplete analysis should not be used as a basis for the radical change in an area-wide policy. The specific conditions of the stream and the watershed. must be reviewed in detail. The Subsidence/Flooding Study supplies extensive knowledge and technical background in) make decisions regarding the relative threat that differential subsidence may have on the depth and extent of flooding. Conclusion is very aware of the effects subsidence can have on flooding. Given the extremes of differential subsidence that are required to produce a marked change in the depth and extent of flooding" and the success of the groundwater control plan, there is no cause for concern at this time (or the foreseeable future) to warrant the drastic change in policy interpretation being proposed by FEMA. ISSUE 3: DUE TO RELATIVELY UNIFORM SUBSIDENCE, IT IS NOT NECESSARY TO USE CURRENT AND ACCURATE GROUND ELEVATIONS IN INLAND AREAS SINCE THE REFERENCE MARKS (BENCHMARKS) SUBSIDE RELATIVE TO THE FLOOD ELEVATIONS The discussion of Issue 2 centered around the effects of subsidence on depth, of flooding for inland riverine flooding conditions. The argument has been presented that for relatively minor amounts of differential subsidence, the depth and extent of flooding will not change significantly. The focus now is on the requirement to use "current and accurate ground elevations," as well as the requirement to convert to the North American Vertical Datum (NAVD). 16 With the change in depth and extent of flooding not being a major issue, the administration of a responsible flood plain management and permitting program must be reinforced in the face of ongoing regional subsidence. Since it has been shown that the depth and extent of flooding for relatively uniform subsidence conditions does not change, it follows that the Base Flood Elevation (BFE) also subsides with the ground (in inland riverine situations). Due to the regional nature of the subsidence phenomenon and the extensive coverage of Reference Marks (RM's or benchmarks), benchmarks within reasonable proximity of the stream will maintain their relative elevation differential: The benchmarks subside along 'with the BFE. So long as benchmarks are not brought in from a long distance, the benchmarks will subside at the same rate as the BFE, allowing the use of the Reference Mark elevations in the effective ?18 with the effective BFE's (see schematic). d1 SB H2 JR ?n Relative Subsidence of Reference Marks and BFE For a Typical Cross-Section 17 In the caSe of Harris County's administration of the building permit program, "shopping" for favorable benchmarks by unscrupulous individuals is prevented. The County Engineer performs all surveys for structure elevations. If conflicting information is found at any RM, the most conservativa elevation is used (that which produces the highest structure elevation). There is also a requirement to set the structure elevation one foot above the BFE. The County Engineer and do not allow encroachments into the flood plain that will increase the BFE (conveyance considerations), and no net loss of flood plain storage is allowed (discharge considerations). Givenr these restrictions, the requirement to elevate structures one foot above the BFE translates into a pure hedge against the inherent uncertainty of predicting BFE's, seasonal changes in channel capacity, effects of subsidence, unintended cumulative effects of individual projects, changes in technical information or methods, minor errors in calculating etc. There is no need to use current and accurate elevations in inland areas if the relative elevations are consistent. The requirement to use current subsided ground elevations with unsubsided BFE's is not supportable technically given everything known about the subsidence/flooding relationship in Harris County. It is also not supportable to require new hydrologic and hydraulic studies for streams if relatively uniform subsidence has occurred. With ongoing (minor) subsidence occurring during and after studies, current ground elevations will always be out of with the study results. This would also negate the need to convert to the NAVD. In coastal areas it is an entirely different story. Under the coastal subsidence scenario the depth and extent of permanent inundation relative to sea level, and increased depth and extent of flooding under tropical storm surge conditions is a ?real threat. A logical solution for coastal areas is to require at least the use of current ground elevations. Prudent management would also use the projection of subsidence under the H-GCSD subsidence control plan (Observed subsidence has essentially been nil for the last 5 years: see Figure 1), as well as the requirement to elevate an additional one foot. In all cases, a registered professional engineer should be required to evaluate the subsidence conditions of the area and certify/seal their findings and recommendations. SUMMARY AND RECOMMENDATIONS It is not technically feasible to expect isolated public works projects, flood control projects, or land development activities to update isolated portions of the FIS models. It simply will not work because the regional nature of subsidence affects the entire stream system. The entire system's hydrology and 18 hydraulics would have to be updated. Regardless of the practical and technical considerations in updating the FIS models, it is not generally necessary given the significant amount of differential subsidence necessary to affect flooding. The successful implementation of groundwater controls has greatly reduced the threat of significant overall or differential subsidence from being created. The desire to update the models for subsidence and to convert to the NAVD can both be addressed when the entire county FIS needs to be updated. The useful life of the FIS is affected by a number issues, such as subsidence, changes in the REC program, calibration to more observed flood data, existing errors in data or calculations, natural changes in channel conveyance, possible conversion to the metric system, etc. The update should be considered for the future whereby the scope and financing of the effort can be planned for by the appropriate agencies. In the meantime, policy and procedures should be documented that will allow the continued processing of the many projects that are generated from this region. The policy and procedures must rely on a reasonable application of the knowledge gained from the relationships established -in the Subsidence/Flooding study, watershed-specific conditions, and the past eight years of successful practical applications. When considerimg if subsidence is relevant, the engineer must review the specific information available for the watershed in question: Is the area considered coastal, transition, or inland? If the area is coastal (or may be due to the effects of subsidence), use current ground elevations with the effective BFE to set building requirements. It would also be prudent to add any projected subsidence for the area based on H-GCSD projections plus one foot. The transition area from coastal to inland can be determined in advance by mapping the increased area of inundation based on actual and projected subsidence relative in; the tidal surge information. Treat the transition area as coastal. If the area is inland, consider the differential subsidence that has occurred over the entire affected reach of the stream. Determine the total differential and the percent change in gradient over the available gradient. If the subsidence is uniform, there is no change in the depth or extent of flooding. Use the effective Reference Marks with the effective to set building requirements. 19 If there is differential subsidence over the streams length, consideration must be given to the relative severity of the differential. Differentials on the order of several feet should be considered carefully, especially if the change in gradient is greater than 20 or 30 percent. If the differential and percent change :h1 gradient are reasonable, the effective Reference Marks with the effective can be used to set building requirements. In all cases, a registered professional engineer should be required to evaluate the subsidence conditions of the area and certify/seal their findings and recommendations. REFERENCES 1) Study of The Relationship Between Subsidence and Flooding." (1986). Turner Collie Braden, Inc., Pate Engineers, Inc., Winslow Associates, Inc. Fort Bend County Drainage District, Harris County Flood Control District, Harris-Galveston Coastal Subsidence District, City of Houston, Houston, Tex. 2) "Groundwater Report, Year Ending December 31, 1994." (1995) Harris-Galveston Coastal Subsidence District, Friendswood, Tex. 3) R.K., and Coplin, L.S. (1990). "Land-Surface Subsidence Resulting From Ground?Water Withdrawals in the Houston-Galveston Region, Texas, Through 1987." Report of Investigations No. 90-01. U.S. Geological Survey in cooperation with the Harris-Galveston Coastal Subsidence District, Friendswood, Tex. 95MDT028.DOC 20 q: . mm uI 0mm sum 56416 (immune-Ila moans) Earn-S911 r0] 9 f; mm"; Ito?ss-sa-rl ID ID (waging) [co? mammal; zzc?cz-srn pt; :moumuqq Mn "no Izr?zrss-n In] on mun?:qu (on-1 anon-ng uni 1535 NI I ?001601 uotsoNID-umsno" 'lelmzm ooqmsmi lo qucdwoo pomsum thymus mlduo 1 unmun-mung plug on animal"; su-u-sa-n in] Hop-mogul?; (Manon soo-m-ss?no:qu (mu-3 no- 5 ummpr pal mg'l set-an my ?"33 .223 umuow 3 I?l El 21133:! Hi HOELDVUWW REFERENCE 2 FIGURE 1 FIGURE 2 I REFERENCE 2 I couu? a ft?nnr new CMIIV I ivemu" I. 1' "wt llnuslnn . I VI -un1 {a . L. .. la": ?mu Lam Nulmusl ,I-Ean F) 1m 3 whim-v.3 l?uadeni . "NI-Jamsou?uuia. amzom amasns chFlgun 7. Map allowing locallon ol oxlensomaler silos, Houslon?Gahte?on legion, Texas MAXIMUM CHANGE IN DEPTH OF FLOODING (FEET) WW THE 3 amnes Tn: vecuAses To 0.2.. 6.0 6 09$ (9?0 06% 4.5 - eeVngQ a \90 Q60 6?f6 4 ?4 a ?0000? 0? 99 0:23.? 9 3.0 A FLOO was me I I can? 0 CREAS 95 2. Foo ?pk EVANS . ?-o.5 1.5 _34.0 5.0 148.0 12.0 16.0 20.0 WITHIN WATERSHED IFEETI (sea 1) NOTE: 1. LEGEND SIMS BAYOU BUFFALO BAYOU BRAYS BAYOU HORSEPEN CREEK LANGHAM CREEK SOUTH MAYDE CREEK BEAR CREEK MASON CREEK WILLOW FORK ill THE SUBSIDENCE VALUES WERE DETERMINED AS THE MAXIMUM LIMITS OF STUDY FOR EACH WATERSHED. DECREASES IN DEPTH OF FLOODING RESULTED FROM THE SIMPLIFIED CASES OF STE EPENING CHANNEL SLOPE. REFERENCE TABLE "-12 FOR DEFINITION OF PLOTTING POINTS BY SUBSIDENCE CASE. A STUDY OF THE RELATIONSHIP BETWEEN SUBSIDENCE AND FLOODING MAXIMUM CHANGE IN DEPTH OF FLOODING vs. SUBSIDENCE HOG-YEAR EVENT) DECEMBER 1986 EXHIBIT lI-12 REFERENCE 1 FIGURE 3 1.2 1.0 AVERAGE CHANGE IN DEPTH OF FLOODING (FEET) 0.5 IOU-YEAR 0.0004 ?-50 >>"30z ?25 .1 n. I 1 ~42. CHANGE IN GRADIENT (PERCENT) I LEGEND 1.0 - I AVERAGE CHANGE IN DEPTH OF FLOODING (FEET0.000480 NATURAL CHANNEL RECTIFIED CHANNEL AVERAGE STREAM SLOPE BRAYS BAYOU BUFFALO BAYOU SIMS BAYOU BEAR CREEK WILLOW FORK (NATURAL) WILLOW FORK MOO-YR DESIGN CHANNEL) LANGHAM CREEK HORSEPEN CREEK MASON CREEK SOUTH MAYDE CREEK 10-YEAR 0.00040 CHANGE IN GRADIENT (PERCENT) A STUDY OF THE RELATIONSHIP BETWEEN SUBSIDENCE AND FLOODING AVERAGE CHANGE IN DEPTH OF FLOODING VS. PERCENT CHANGE IN GRADIENT QaEcEMgefa EXHIBIT II-11 REFERENCE 1 FIGURE 4 35"11' gunning?, mansguquh?ailer 4-964 Human3-5 Awtoxim?niv ?6 mi" ?0 13'? L'V'm'p.90. . I o] .t'i I {747ypreffn- ?1 Inks LIBERIY - It- r? "r 5 CHAMBERS ?fr. i . I v" State 22-9brook . - - - Jahnw? gr I a? Space i .1 If" I 1 - 1.51 .5 r/ 5 {7?grFigure ?5-"ADDroximate land?surface subsidence. 1978-87. EXPLANATION 1.5 OF EQUAL LAND-SURFACE SUBSIDENCE- - Dashed where appronmatcly located. Interval 0.25 and (15 leet 20 MILES manners? DIRECTION OF FLOW AND RELATIVE SIZE OF PRIMARY DRAINAGE BASINS NOT MAPPED FOR 1973-5133 1987-1995 NOT AVAILABLE, YET COASTAL SUBSIDENCE SPECIAL FLOOD HAZARD AREAS FIGURE 5 MAP SOURCE: REFERENCE 3 REFERENCE 3