Water and Productivity Impacts for the NBDC

Water and Productivity Impacts for the NBDC

Charlotte MacAlister Birhanu Zemadim

Teklu ErkossaAmare Haileslassie

Dan FukaTammo SteenhuisSolomon Seyoum

Holger HoffKinde Getnet

Nancy Johnson

Nile Basin Development ChallengeScience and Reflection WorkshopAddis Ababa, 4-6 May 2011

After measuring and acquiring time series data:-Next step??

• Hydrological Modelling using SWAT/SWAT-WB – Simulating the different physical processes in both time and space with

reasonable accuracy– Determine water use and productivity at different locations in the landscape (e.g.

from trees, cultivated areas and pasture etc.)

• Using the simulated model evaluate the possible implications for hydrological fluxes of different scenarios of RMS interventions

e.g. the installation of ponds and tanks, bunds or terracing etc.

• Finally determine the impact of up-scaling RWM interventions on downstream flows as well as water productivity at different locations in the landscape.

Field monitoring

Hydro-Meteorological Monitoring in the Study Landscapes

Objectives1. Developing primary data to understand water budgets, fluxes and

flow pathways- Hydrological processes. 2. Estimating water-use and water productivity within the framework of

Integrated Rainwater Management Strategies, Policies and Institutions.

Data Obtained will be used to calibrate and validate computer models (e.g. SWAT) which can then be used to simulate:

• Hydrological Processes

• Investigate the possible hydrological implications of RMS interventions

Equipments to be Installed and Variables to be Monitored

• Automatic Weather stations measuring rainfall, temperature, relative humidity, wind speed and wind direction, solar and net radiation and soil temperature. One in each catchment

• Manual Rain Gauges distributed across altitude and space. 5 in each catchment

• Pressure Transducers 1. Stream flow measurement at catchment outlet to measure river stage (converted to flow using a rating equation, determined from current meter measurements). One in each catchment 2. Groundwater level measurement Five in each catchment

• 20 Stage Boards at the catchment outlet to enable manual measurement of stage (converted to flow using a rating equation, determined from current meter measurements)

• 3 Soil moisture profile inside the catchment to measure volumetric moisture content

• Dip Meters portable, reliable instrument for measuring the water level and total depth in boreholes, wells etc.

• Sediment Sampler Done manually

Equipment to be installed and corresponding variables to be monitored

Challenges:

- Limited recording of parameters e.g. ET creates difficulties for scaling/projecting both at different altitudes and scales

- No actual recording of impacts of RMS

- Short duration of monitoring period

Remote monitoring:the solution to upscaling (hydromet) parameters?

GEONETCASTNRT global network of satellite-based DDS

(space-borne, air-borne, in situ data)

Source: EUMETSAT brochure

Uses of GEONETCAST:

• Filling precipitation data gaps

• Measurement of 30 minute rainfall used in erosivity estimation

• Aerial rainfall distribution

• Soil moisture estimation

SEBS input

LST Albedo NDVI LAISZA DSSF DEM

SAF SAF VGT4Africa

VPI

SAF MSG2 SAF SAF SRTM

15mts 1day 10day 1day Calculated 30mts 1day once

Resampling by 2000m georeference

SEBS = estimating atmospheric turbulent fluxes and

surface evaporative fraction using EO data in the

visible, NIR and TIR

SWAT 2005, User Manual

•I30, Rday and imax can be derived from the 15’ rain intensity time series

APPLICATIONS cont’d…

SWAT 2005, User Manual

Disturbance Index

Challenges:

• Continuous validation is required (ground truthing)

• Data storage issues

Water productivity of mixed crop-livestock systems:

progresses and challenges

High unproductive water losses

Example: Runoff + EvaporationHigh nutrient depletion

WP scenario ought to change!

-180

-160

-140

-120

-100

-80

-60

-40

-20

0

20

40

N P K

Bala

nces

(kg

ha-1

yr-

1)

RichMediumPoor

Full nutrient balances on Teff

Blue Nile Highlands: >95% rain fed and major sources of livelihoods

Crop WP gaps are enormous. Implication for LWP

Objective:

Develop strategies and menu of practices for different recommendation domains and understand the impacts of the strategies for water productivity.

WaterDepleted

output alAgricultur ty productiviWater

How we do: Understanding the system

HH survey(30-40 farms per

system)

Focus farmers:(landscape positions

resources flow monitoring)

Land

scap

e/w

ater

shed

bou

ndar

y

• Feeing and feed sourcing;

• Livestock products;• Livestock services• ME demand

•FU factor•Agricultural water Partitioning

Water flow and biomass H2O

productivity: by land use and crop type

Cropping pattern

Monitor:•Weather and H2O flow •Nutrient flow •Crop management and performance

Crop WP

How we do: Understanding the system

Crop WPCrop WP

Crop

Weather

Modeling Crop WP:Current VS RWMS

Potential Models: AquaCrop, APSIM

Soil

Management: Current Vs Improved Scenarios

Map + Management practices (Students)

Selected field sample + Sediment quality (students)

Review + Field survey

RCM +weather stations

Out scaling to basin scale

For Crop WP:• Identify similar units

represented by farming systems of the study landscapes

• Acquire baseline data for the identified systems

• Run models for the systems

• Integrate system outputs into basin scale

Challenges

• Resources demanding! Some tools (e.g. APSIM) are data intensive

• Whether biophysical factors such as soil, climate, landscape and crop types … details are considered in the similarity analysis for basin wide scaling up needs clarification

• Capacity of soil laboratories to handle runoff samples

• Data for large scale base line (soil, weather, crop management practices, livestock population)

Green - Blue Water

Green-blue water baseline – framing the NBDC

assessing: • green-blue water availability &• variability• storage (buffer capacity)• agricultural productivity

1. CP basins2. Nile sub-basins3. Blue Nile sub-basins

mapping spatial patterns, hotspots, opportunities

Nileblue water: ~ 1300 m3 cap-1 yr-1, green ~ 900(crop) water productivity: ~ 700 kcal m-3

potential kcal production: 4300 kcal cap-1 day-1

in dry years (10th percentile): 3900 kcalIn 2050: 1900 / 1600 kcalBUT: huge spatial variability along the Nile LPJ-based

in a nested approach

(„difficult hydrology“ – David Grey)

rainfall variability and drought risk(ILRI)

blue per capita 2000 / 2050

green per capita 2000 / 2050

CoV & 10th percentilegreen - blue

water productivity

Nile (framing Blue Nile for taking into account downstream consequences)

IDIS -> NBI subbasins

water availability

1) CPWF, Mulligan2) IWMI, Karimi

CPWF and IIASA data

additional data:

integration with socio-economic data, e.g. market access (JRC):

blue water storage from „rethinking water storage“ project

yield gaps from Licker et al (Wisconsin)

Blue Nile

Blue Nile SWIM model: focus on water productivity & climate change

green-blue water availability at sub-basin scale?

Effects of interventions / SWC

e.g. rainwater harvesting

and consolidated with other evidence from the Blue Nile

starting from WOCAT database

and

suitability mapping (Dile 2010)

can local WOCAT experiencebe generalized / upscaled

Lake Tana SWIM model for simulating the effects of outscaling / upscaling

for assessing landscape scale effects of increasing adoption?

Resilience (landscape) approach

„social ecological systems“

integrating water fluxes and productivitieswith institutions and their position / effectiveness re uptake of measures

node-based WEAP node-based social networks

Challenges:

•Lack of data at smaller scale

Hydrological Modelling with SWAT

Runoff plots

Maybar)

16 37 43 64 slope of land

Runoff Coefficients

Hydrological processes + impacts

Surface runoff decreases with steepness

Hydrological processes + impacts

Hill slope Areas

Surface runoff

infiltration

interflowv

Runoff and Scaling issues in SWAT

Current limitations:– Limited scaling abilities due

coarseness of soils and land use data base

– Can upscale by grouping HRUs, but no ability to downscale above the resolution of the finest initialization dataset.

– Predicts infiltration excess runoff instead of saturation excess

– Topographic data not used that is driving force for hydrology

SWAT-WB initialization method allows continuous scaling to model full basin to sub-catchment

scale phenomenon.

Allows

TI/STI based

Predicts overland flow locations correctly

Down-scaling to DEM Resolution

Up-scaling with physical attribute instead of judgment

Challenges:

•Current data availability – for calibration and validation•How to simulate actual RMS within the model

Evaluating Water Resources: Upscaling Impacts of RMS

Objectives of Water Resources Modeling (WEAP)

1. Assess integrated impacts of rainwater management interventions and large-scale development projects on water availability, sediment load (and groundwater recharge)

2. Device mechanisms of incorporating rainwater management interventions in IWRM planning

3. Investigate up- and out-scaling options of rainwater management interventions using sustainability criteria of satisfying downstream flow requirements

Methodologies and Approaches

General Approaches:

• Represent RMS interventions scenarios in SWAT (parameters and management options)

• Perturb inflows to large dams and river junctions according SWAT simulation of RMS interventions

• Evaluate the bio-physical sustainability criteria for large-scale and RMS scenarios

Potential Scenarios:1. Baseline (current) 2. Medium-term3. Medium-term + RMS 4. Long-term5. Long-term + RMS

WEAP Schematization

Dams / Reservoirs

Irrigation demands

River Networks

Modeling FrameworkW

ate

rsh

ed

sS

ub

-ba

sin

sA

ba

y B

as

in

SWAT Model

ClimateSoilLU

WEAP Model

RMS Scenarios

L-SDev’t

Scenarios

ImpactEvaluation

Large dams and

River Junctions

Challenges• The WEAP reservoir and

river junction nodes should match with SWAT sub-basins

• WEAP doesn't have sediment routing capability

• Aggregated RMS scenarios are not yet defined (targeting)

Translating hydrology and productivity into environmental, social and economic

impacts

WHAT is evaluated?

Current practice (crops, livestock, trees, management practices) is compared to selected alternatives in terms of their impacts on:

• Farm profit (economic benefit)

• Employment opportunities (social benefit)

• Environmental externalities (ecosystem benefit)

WHY multicriteria analysis?

• To know the possible outcome in terms of profit, employment, and externality

• To understand trade-offs between criteria and between upstream and downstream land and water users

• To estimate values for environmental services that could be the basis for compensation schemes

HOW to estimate impact? Linkages in N4

• Bio-physical and hydrological characterization of watersheds and prioritization of areas for interventions (SWAT & WEAP)

• Identify RWM practices and how they would be incorporated into land use systems

• Gather data on productivity and environmental impacts

• Gather data on socio-economic costs and benefits

• Optimization model to get economic, employment and environmental impacts of RWMS scenarios

.Biophysical data

Hydrologic modeling

Externalities

OptimizationSocio-

economic data

RWM practices

Some examples on data inputs

• Externalities

• Socio-economic issues

• Erosion and sediments through land use

• NPK release to water resources• Additional water available

• Land (allocation and constraints)• Crops, forages, livestock, forest

(production) • Costs and prices for crops,

livestock, forests• Management practices and costs• Labor (costs and incomes)

Scaling out impacts - is basin level estimation appropriate?

• Run model for representative sites

• Adopt results of similarity analysis

• Scale out model results of representative sites (to similar sites) at basin level

• Aggregate the scaled-out results

Challenges and limitations

• Characterizing and prioritizing scenarios with one or more RWMS

• Scaling out impacts to a basin level

• Incorporating gender and equity issues

• How to validate the model outputs?

Challenges Summarized:• How do we deal with data gaps at field and basin

scale; validating the model outputs? (Dan + Birhanu)

• How do we characterize/ aggregate/prioritize RMS scenarios – especially when we don’t have actual data measuring the impact of any specific RMS? (Solomon + Amare)

• What alternatives exist to address scaling out RMS practices and process/impacts to a basin level? (Tammo +Teklu)

• What alternative options do we have to link biophysical and livelihoods issues? (Kinde + Nancy)