Escolar Documentos
Profissional Documentos
Cultura Documentos
STAKEHOLDER WORKSHOP
11th February 2005
PROCEEDINGS
Hydrology and Water Resources Okavango Delta Management Plan
List of Contents
1 BACKGROUND
2 PROGRAMME
3 PRESENTATIONS AND DISCUSSION
3.1 Introductions
3.2 Integrated Hydrologic Model
3.2.1 Introduction
3.2.2 Conceptual Model
3.2.3 Evapotranspiration
3.2.4 Surface Waters
3.2.5 SubSurface Waters
3.2.6 Outputs
4 SCENARIOS
4.1 Introduction
4.2 Hydrologic Inputs
4.3 Presentation of Results
4.4 Baseline
4.5 Upstream Water Resources Developments
4.5.1 Introduction
4.5.2 Dams in Angola
4.5.3 Irrigation
4.6 Abstractions from Delta
4.7 Climate Change
4.8 Management Planning
5 DISCUSSION
5.1 Introduction
5.2 Database Management
5.3 Integrated Hydrologic Model
5.4 Scenarios
5.5 Presentation of Results
5.6 Stakeholders’ Responses
6 CONCLUSIONS
6.1 Summary
6.2 Recommendations
1. INTRODUCTION
1.1 Background
The Hydrology and Water Resources is one of the twelve components comprising the
Okavango Delta Management Plan project. The project and the component
commenced in May 2003.
The development objective of the Okavango Delta Management Plan is integrated
resource management for the Okavango Delta that will ensure its long term
conservation, and that will provide benefits for the present and future well-being of the
people, through sustainable use of its natural resources.
In line with this development objective, the immediate objectives for the Hydrology
and Water Resources component are improved water resources planning, and
monitoring and evaluation in the Okavango Delta, based on an enhanced capacity of
the Department of Water Affairs. Corresponding to these objectives, the outputs of
the component are:
(1) A comprehensive quality controlled database comprising existing climatic,
hydrologic, surface water, ground water and sediment data for the
Okavango Delta.
(2) Recommendations on the improvement and expansion of the Okavango
Delta monitoring network.
(3) A digital Topographic Model of the delta.
(4) An Integrated Hydrologic Model for the delta.
(5) The analyses of the impacts of water resources scenarios for the Okavango
Delta Management Plan.
(6) Capability within DWA to maintain and operate the Integrated Hydrologic
Model for the establishment and implementation of the ODMP.
While all aspects of the component were discussed, the Workshop focussed on
outputs (4) and (5), the Integrated Hydrologic Model and its application to impacts of
water resources developments in the delta and the basin upstream, with comments
and discussion on the model and the format of the presentation of the outputs.
1.2 Objective
The Hydrology and Water Resources workshop was convened to:
• Inform project partner institutions and the community about the role of
the integrated hydrologic model in management planning
• Afford the project partner institutions and the community the
opportunity to comment on the outputs of the model and to suggest
ways in which the outputs or result can be made usable by them
• Afford the project partner institutions and the community the
opportunity to suggest management scenarios relevant to their
respective institutions
• Ensure that this component is community driven or encompasses the
views of the majority of the stakeholders from the beginning
• Guarantee ownership, commitment and sustainability of the project
• Minimise risk during implementation of entire management planning
1.3 Programme
The programme for the one day Workshop is presented in Annex 1. The first session
covered the background to the establishment and the implementation of the
component in DWA, the recommendations for improved hydrologic monitoring in the
delta and their pending implementation and data collection and management.
This was followed by a description of the Integrated Hydrologic Modelling system, and
its set up to represent the hydrology of the Okavango Delta. The next four
presentations covered the application of the model to represent existing
developments in the delta and the basin upstream, and hydrologic conditions as they
would be in the delta given the implementation of a range of potential water resources
developments. Each presentation was followed by a discussion session.
The team implementing the component comprises DWA staff, and national and
international consultants. An important output from the component is capacity
building in DWA, to maintain and operate the Integrated Hydrologic Model. This
entails technology transfer from the consultants to the DWA staff. The technology
transfer is carried out in a three pronged approach:
• A series of formal training courses in DWA Gaborone on the main
elements of integrated hydrologic modelling
• On-the-job training, whereby the DWA staff learn by doing, setting up
and applying the model under the guidance and supervision of the
consultants
• Overseas training, most likely MSc courses in recognised international
hydrologic institutions
DWA has assigned four permanent staff members to work full time on ODMP. As
part of the technology transfer process, each of these four staff presented an aspect
of the model application in the Workshop.
The presentation of the model and its application was followed by an open discussion
by the participants. Each sector was given a slot to allow the representatives to voice
their particular comments.
The model results at this stage are preliminary, and the focus was more on are the
hydrologic and water resources development data appropriate, and is the
presentation format of the results meaningful and useful to the respective ODMP
sectors, in the context of the Okavango Delta Management Plan.
1.4 Participation
The list of participants is given in Annex 2. Despite an early dispatch of invitation
letters in advance to representatives of the different components, it became apparent
that most components were not represented at the workshop. Most notable
absences were from the technically oriented components such as Sustainable
Livestock Management, Vegetation Resources Management, Waste Management,
etc.
Of the components that were present it appeared that these representatives were not
familiar with the project or are not dealing with the project on a daily basis judging
from their contribution to the discussions on the role of the Hydrology and Water
Resources component. It appears that project partner institutions still have a
combined workload of their routine work and new duties related to ODMP. This
added responsibility is somewhat too great and affects the ability of project officers to
deliver efficiently and effectively.
Each component should identify a suitable person to act as a focal point for liaison
with the Hydrology and Water Resources component, and this person should attend
future such workshops and meetings.
2.2.3 Evapotranspiration
Evapotranspiration from the open water and vegetation in the delta is key to
describing the water balance. Of the main inflows to the delta, river flows from the
basin upstream and rainfall over the delta, upwards of 95% is lost to
evapotranspiration, the remainder infiltrating to ground water and downstream surface
water outflows.
The greater the inflows and the corresponding extent of flooding in any one year, the
greater the losses to the atmosphere. The rate of evapotranspiration is crucial to the
water balance and the extent of flooding. The model uses a soil-vegetation-
atmosphere transfer (SVAT) mechanism to describe the process.
The SVAT component comprises a two layer soil-canopy system linked by a network
of resistances. This enables representation of the latent and sensible heat flux
between the soil, the canopy and the atmosphere based on humidity and temperature
gradients controlled by aerodynamic and stomata resistances, and atmospheric
conditions. The model distinguishes losses from open water, the soil and the
vegetation, and thus has a dynamic coupling with the surface and subsurface waters
(slide 3). The vegetation dynamics are based on remotely sensed data with limited
ground verification (see slides 4 and 5). The relationship between the extent of
surface water and the rate of evapotranspiration is shown for the normally and
seasonally flooded areas in slide 6.
2.2.6 Outputs
The outputs from the model are designed to support broad based management
planning in the Okavango Delta. The outputs are presented in grid or map format
showing the spatial variation in the parameters over the delta, time series (plots of
various parameters against time) and summary tables of results. The outputs may be
summarised as:
• The actual evaporation from open water and the transpiration from
vegetation
• Surface water levels and flows through the rivers and over the flooded
swamp areas, and the extent of flooding
• The moisture content of the soil
• The ground water levels and flows
• The overall water balance, including inflows, outflows and changes in
surface and subsurface storage, for the delta and selected areas of the
delta
• Sediment transport along the main channels of the delta, showing
patterns of erosion and deposition
The Integrated Hydrologic Model does have practical certain limitations, and it is
important that these are recognised by the stakeholders and management planners.
Channel Switching
While the model can predict the impact of cutting and dredging a new channel
through the delta, and the impact of closing an existing channel through the
encroachment of vegetation, it is not able to predict the occurrence of the formation of
new channels, and the realignment of existing channels, which may be initiated by the
movement patterns of large mammals and vegetation growth.
Certain trends in sedimentation and channel development and decay may be
predicted by the application of sediment transport to the model. Work on this area
has just commenced. Detailed analyses of selected reaches and areas of the delta
may be carried out through the application of two dimensional morphological models.
This is beyond the scope of the present component activities.
Physical Phenomena
The model simulates physical hydrological phenomena only. Studies have revealed
in particular the importance of the interdependency of physical and biological
phenomena, for example the movement of large mammals opening new flow paths
through the vegetation, sedimentation and vegetation in closing existing channels.
Hydrologic Output
The hydrologic model is integrated in the sense that it integrates atmospheric, surface
and ground waters. It does not integrate the various sectors of the management plan.
The output of the model describes the dynamics of the waters of the delta, under
present conditions and under future water resources development scenarios. It does
not describe the impact on wildlife, vegetation, livestock, fisheries, tourism, etc.
Impacts on and management criteria for these sectors have to be assessed by their
respective plan components, based on the outputs of the levels and flow patterns
presented by the Hydrology and Water Resources component.
2.3 Scenarios
2.3.1 Introduction
The ODMP Integrated Hydrologic Model has been run to simulate present
development conditions in the delta, and development conditions as they may be
given a range of water resources development scenarios in the delta and the basin
upstream:
• Upstream water resources developments: dams and irrigation
schemes in Angola and Namibia
• Surface and ground water abstractions from the delta
• Regional climate changes
These scenarios represent possible conditions in the basin, notionally in the year
2025. Each scenario is compared against the present conditions which serve as a
baseline.
The development states are considered static. It would be possible to represent a
dynamic basin condition, as populations and corresponding abstractions increase and
water resources developments are implemented year on year. This is not considered
as it would introduce complications regarding the coincidence of development with
hydrologic events, and the impacts would be more difficult to assess. For example,
the completion of a dam before a sequence of dry years would have a very different
impact if a wet period followed completion.
• The data should reflect how the delta may recover in a period of
normal inflows after a dry period.
The long term inflow in terms of volume per annum from 1933 to date is shown in
slide 14 for both Mukwe in Namibia and Mohembo in Botswana. Whereas at Mukwe
the river is more or less confined to the gauged channel, at Mohembo the river has
broadened out with a wide flood plain. It is not possible to gauge the flows over the
flood plain, and Mohembo will not represent the total discharge. This is most clearly
evinced in the high flow years 1998/99 and 2000/01.
An anomaly appears in the data from around 1950 to 1980, where the total inflow at
Mohembo appears higher than Mukwe. As there is no significant inflow to the river
between the two gauges, this can only be explained by gauging errors. The series
after 1980 is more consistent, with Mukwe showing a higher inflow than Mohembo.
The period from 1992 to 1997 represents the sequence of the five lowest inflow
years, and is therefore chosen to provide the input data to analyse the impacts under
critical conditions, for both the existing state and possible future states with water
resources developments.
The following five year period, from 1997 to 2002 represents relatively normal
hydrologic conditions in the delta, and has been used to assess how the delta may
recover after a sequence of dry years (see slide 15).
A high inflow period could also be selected, eg 1987 to 1992, to assess the impacts
under high inflow conditions. It is not planned to analyse this period as:
• The conditions will not be critical to a management planning exercise
• The number of impacts which have to be assessed in terms of flows,
depths, etc will be further increased
Appropriate time series of inputs (inflows, precipitation, etc) have been prepared and
applied to the existing state of the delta. The model results are used as a baseline
against which future developments are assessed for their impacts. The inputs are
modified to represent conditions which may prevail under future conditions.
To summarise, while the inflows to the delta will be dynamic, ranging over two five
year periods representing critical dry and normal average hydrologic conditions, the
existing and future states will be static. This implies that for the duration of the five
year model run period, the population for instance will not be increasing, no new
dams or irrigation schemes are built, etc.
At present, the Baseline and Scenario Conditions have been simulated for the critical
dry hydrologic input period from October 1992 to October 1997. Summary results are
presented and discussed in the following sections. Slides illustrating the outputs are
presented in Annex 3.
separate zone (slide 16). The probability, delta area and a zone description are given
in the table.
(1) Animated sequence of flooding in the delta showing the monthly variation in
flooded area and depth. (The AVI format file is available from DWA, size
around 110Mb.)
(2) Map based on the one kilometre grid showing the minimum depth of water
in the delta for the five dry years.
(3) Time series showing the variation in the depth of flooding over the five year
period. The depth is averaged for each of five zones (see table).
(4) Map based on the one kilometre grid showing the minimum level of soil
moisture in the root zone (the zone from which plants draw water) in the
delta over the five dry years.
(5) Map based on the one kilometre grid showing the maximum depth to
ground water in the delta over the five dry years.
(6) Summary tables are prepared showing the following:
• Water balance showing inflows, outflows and storage changes
• Baseline conditions and impacts for the surface water, soil
moisture and ground water, for the entire delta and for each of
the five zones
A wide range of outputs of all hydrologic parameters in various formats is possible,
and can be tailored to the needs of the individual stakeholders. It is possible also to
select different zones to calculate average time series and summary results. The
above selection has been made by the component as samples of what may be useful
and meaningful to the stakeholders, and to generate discussion.
2.4 Baseline
Surface Water
The animated depth of flow sequence (slide 17) shows the expansion and contraction
of the flooded area and depth as the sequence of flood waves for each of the five
years propagates through the Panhandle and spreads out through the delta. Slide 18
shows the detail for the lower envelope of flooding for the upper delta. Each square
is one square kilometre.
Time series of the depth of surface water are shown for each of the five zones in slide
19. The Panhandle is the first area affected by the flood wave from upstream and,
being relatively confined within a flood plain around 20km wide, has the maximum
range from the seasonal wet to dry periods, around 0.6m through the five dry years.
As the flood wave propagates through the delta, the range in water levels declines.
The accumulated impact of the drought years is also felt more as there is a general
decline in the flood depth.
As the summary table (slide 20) shows, the area flooded ranges from 1,513km2 to
12,077km2, while the depth ranges from 0 to around 5m. The average depth in the
normally flooded area ranges from 0.11m to 0.92m.
SubSurface Water
The distribution of the soil moisture content is shown as a deficit below saturated
level, as the lower envelope (maximum deficit) for the five year period in slide 21.
The average deficit ranges from 0.56m in the rarely flooded area, to 0.12m in the
normally flooded area.
The depth to ground water over the delta area is shown for the lower envelope
(maximum depth) over the five year period in slide 22. The depth ranges from an
average of 1.88m in the rarely flooded area to 0.37m in the normally flooded area.
Water Balance
The overall water balance for the Baseline and Scenarios is shown in slide 23.
Negative values indicate an inflow to the delta (eg precipitation and river inflow), while
positive values indicate an outflow (eg evapotranspiration and river outflows). The
Total should be zero: the small value indicates a very minor continuity error. The
negative overland and subsurface storage changes indicate a decline in storage in
these levels.
redistribution of the natural hydrograph: flood flows are reduced while dry season
flows are increased. (The impact appears only to be significant in 1993 and 1996 –
this needs further investigation.)
The impact of the dams on the surface and subsurface waters of the delta is minimal,
given the limited apparent impact in three out of the five dry years, and the fact that
there is no nett water loss in the basin upstream.
2.5.3 Irrigation
The impact of the potential irrigation scenario on the inflow to the delta is shown in
slide 25. The impact is greatest in the dry period, when water is withdrawn for the
relatively large area of winter crops in Angola. The summer withdrawals for irrigation
in Namibia are significantly less.
Correspondingly, the impact on the delta in the flood period is minimal, while in the
dry period the overland depth of flow is reduced, particularly in the Panhandle and
normally flooded areas, and the minimum flood extent is reduced by 660km2, or 40%
of the present dry period area (slide 26).
1
http://www.met-office.gov.uk/research/hadleycentre/models/HadCM3.html
3. DISCUSSION
3.1 Introduction
The following sections record the comments made on the foregoing presentations,
and the responses given by the hydrology team (in italics).
3.4 Scenarios
(1) You have run the scenarios based on dry period(s). Could you please
revise this? You might be missing out on recharge to groundwater as a
result of high floods. A single high flood could recharge an aquifer for the
next twenty years, for example.
It is the intention to run the model for a period of normal flows, to see how
the delta recovers from a drought. As suggested, the model will also be run
for high floods (this is also important for sediment transport in the delta.)
(2) In taking the dry period as critical, you have to consider floods as also
critical. Floods need to be covered as well.
The delta is not under stress during high flow periods, and the impacts of
water resources developments will not be so keenly felt and therefore not
critical to management planning. See also response to (1) above.
(3) People are more worried about the switching of channels. Maybe this
aspect needs to come out in your model outputs as well.
The introduction of sediment transport to the model (which has just
commenced) will give an indication of possible future switches, and
redistribution of delta flows. Switching and realignment of channels has a
chaotic component (the butterfly effect on global climate is akin to a
hippopotamus going walkabout in the delta), and accurate predictions are
simply not possible.
(4) Looking at the model’s level of resolution, critical management issues will
not be answered.
The component has been conceived and ToR written (in 2001/02) to
address the broad scale management planning issues (water resources
developments upstream, climate change, abstractions from the delta, reed
cutting and dredging). On this basis the model approach has been set out,
and the model set up and applied. Preliminary applications suggest that
the main threats to the delta will arise from upstream irrigation and climate
change. These are the critical management issues related to hydrology.
(5) Channel blockages impacts need to be covered in detail. Channel switching
is another fundamental management issue.
The model can simulate the impact of channel blockages and channel
switching. As discussed in (3) above the model cannot predict these
occurrences.
(6) You mentioned that your model could handle sediment transport. Do you
have something on sediment transport now?
An approach to introducing sediment transport to the model has been
drafted (pending report on the Integrated Hydrologic Model).
(7) You seem not to be representing your conditions.
See response to (6) above.
Fisheries
(1) Water quality is very critical as this have impact on fish population
dynamics. Sedimentation is also critical as it has a bearing on fish
populations.
Sedimentation is indeed critical to fish populations, as fine organic rich
sediment is washed out from the flood plain into the channels by the flood
waters, resulting in oxygen depletion.
The MIKE SHE integrated hydrologic modelling system has the capability to
simulate water quality. This is not within the present scope of ODMP.
Systematic data (including DO and SOD) do not exist at present on which
to develop such a model. Improved monitoring recommendations include
the introduction of systematic sampling and analysis at the key inflow
station of Mohembo, and monthly in-situ measurements of salinity and pH
at around 20 discharge stations throughout the delta.
Water Affairs
(1) Is MIKE SHE an instrument or what?
MIKE SHE is one of a suite of water and environmental software packages
developed and marketed commercially by DHI Water and Environment.
The MIKE SHE package is the most advanced generic (can be applied to
any system) hydrologic software available. MIKE SHE has a consistent
development history over thirty years, and has been applied to hundreds of
hydrologic systems worldwide.
(2) Can the model predict channel shifts?
The sediment transport model will predict erosion and deposition leading to
channel aggradation and eventual closure. The delta morphology is
inherently unstable and chaotic, and no model can predict channel
abandonment with higher certainty than what is constrained by natural
variability (chaos). Nonetheless, the impact of major channel shifts can be
predicted, as well as possible management measures to mitigate such
shifts.
(3) Mr Dikgomo mentioned that Hydrology and Water Resources component is
one of the twelve. What are the other eleven components?
The CTA may best respond to this question, though it is suggested that
representatives from participating organisations take the trouble to inform
themselves of the ODMP structure, objectives and outputs before attending
such fora.
Land Board
(1) Is it possible for the model to give us depth of water table within at least
500 metres from the delta? This can help us with land use. This will also
help us determine how far we can avoid pollution in land allocations.
The spatial resolution of the model is one square kilometre. The model can
give the various hydrologic parameters such a depth to ground water within
the general area (within a few square kilometres). This should be quite
adequate for the purpose of broad based management planning.
(2) Can the model add value to Tawana land board land use and management
plan (2004-2029)?
The Hydrology and Water Resources component is concerned with
hydrology and water resources. The outputs from the component are
hydrologic data and predictions for the delta. It is up to the various sectors
and components of the management plan to assess the impact of these
hydrologic parameters for their own purposes. The H&WR component can
assist in the interpretation of the hydrologic data.
Tourism
(1) There are lodges and camps located close to the river channels. Can the
model be able to help on the implications of pollution?
The model can simulate the pattern of flow in the delta, which may give a
useful indication of the movement of pollution, though not the output which
may be given by a water quality model (advection, dispersion, decay, etc).
General
(1) This component (Hydrology and Water Resources) is so detached from the
rest. It should relocate to Maun.
The component is located in Gaborone as the capital has better
infrastructure. Progress on the component would be constrained by the
limited infrastructure in Maun.
(2) To what extent is your data real time series that it could be relied upon?
The data available from the delta at present are poor, largely owing to
difficulties in access, and disturbance of gauging stations by humans and
large mammalian wildlife. Lack of systematic use of the data to date have
also contributed to neglect. Improved monitoring is necessary for
management planning, and an improved hydrologic model with finer
resolution and greater accuracy. Recommendations have been prepared
for DWA to improve hydrologic monitoring, and it is hoped these can be
implemented in the course of 2005.
(3) The presentations are too technical. When will ODMP reach out to us?
Someone mentioned that communication and or/ information dissemination
is a problem. Why cant we find other Batswana who could do the job to
convey the information in an audience screened and targeted way?
The hydrology of the delta is complex, and requires a complex model for a
realistic simulation of the processes. The presentations were intended to
give all participants an understanding that the model is indeed capable of
simulating the key hydrologic phenomena of the delta. It is hoped that this
has been achieved, though it is recognised that the depth of understanding
4. CONCLUSIONS
4.1 Summary
Prof Francis Sefe from the consultants’ team summarised the Workshop discussion
as follows:
(1) There were discussions on both spatial and temporal resolutions of the
model, which need clarification.
(2) It was observed that data quality and availability is a constraint to accurate
modelling of the hydrology of the delta, and providing physical evidence to
back the model conclusions.
(3) In the selection of scenarios, it is critical to understand that the individual
scenarios are not stand alone.
(4) The choice of what constitutes the baseline scenario was discussed. It has
to be emphasised that the inclusion of exceptional floods is also important
and hence not only the dry period should be used.
(5) The presentation was multifaceted.
(6) People could not easily locate any landmark features on the map. This
need not be. Villages could be labelled, and even the rivers themselves.
(7) Though water quality is not in the scope of the current modelling, water
quality is very important and this has to be looked into in the future.
(8) Ground water fluctuations at the periphery of the delta need to be looked
into.
(9) Various stakeholders have difficulty interpreting the results. This is a critical
issue that needs to be addressed. The target audience should reasonably
be able to assimilate the information relayed to them.
4.2 Recommendations
The following recommendations are made with regard to the future interaction
between the Hydrology and Water Resources component and the other ODMP
participants.
(1) While the introductions to the set up and application of the Integrated
Hydrologic Model attempted to convey the complexity of the delta
hydrology and its computer based representation, further effort needs to be
made to present the model and its results in a manner which is fully
comprehensible to the Stakeholders, most of whom do not have in depth
hydrologic knowledge.
(2) It was originally believed that presentation of the model to Stakeholders in a
combined forum would allow the interchange of ideas and knowledge
among the Stakeholders as well as with the Hydrology and Water
Resources component. The resulting discussion tended to be dominated
by those with technical hydrologic knowledge at the expense of those
without. The next round of discussions should be on a bilateral basis, such
that the discussions can proceed at a level appropriate for the individual
Stakeholders, and their needs in respect of hydrology and water resources.
(3) Each ODMP component should identify a suitable person to act as a focal
point for liaison with the Hydrology and Water Resources component. This
person would attend all future meetings and Workshops aimed at
integrating hydrology into ODMP, thereby providing continuity to the
process.
(4) The Data Management component of ODMP implemented in HOORC is
crucial to integrating and interpreting information among the partners. A
member of the Data Management component should also be involved in
discussions to demonstrate how this can be realised in practice.
ANNEX 1
ANNEX 2
WORKSHOP PARTICIPANTS
Name Surname Organisation Address Telephone E-mail
ANNEX 3