JCP Metamodel

Expert reflection session (Water & Climate Change) July 16, 2020 Code of conduct

• Please mute your microphone, if you don’t talk • Ask your questions in chat (will try after the topic presentations) • Raise your hand during discussions in Teams, if you want to make a remark 3 Program

• Opening and introductions by chair Abu Saleh Khan (Executive Director, IWM) • Presentations progress Bangladesh Metamodel • Introduction (Marnix) • Detailed concepts/assumptions of WBM and NM (Riasat) • Calibration results of NM (Shahadat) • Calibration results of WBM (Kymo) • Results of runs with MM (Kymo) • Surface water salinity (Shahadat) • Closing by the chair • General reflection and recommendations Example of use of Metamodel for an investment project

Agencies GED

(Re-) (Re-)select Explore projects on SDG formulate clusters of & BDP goal contribution, project projects impacts and investment

Assess & Assess (Re-)select projects into compare impact programs impact

Concept Sector Assess and compare note Action Plan impact

Impact explorer BDP Investment Plan Input to 8th & 9th 5YP & Impact explorer & Annual Development Plans Program manager Metamodel modules

Metamodel engine module workflow State Indicators Decision Support Indicators Environmental flow (m3/s)* Annual rainfall damage (Taka) Network Salinity Dry season river flow (m3/s) (water in River navigability (km/class)* river) Annual flood extent (km2) Rural access to safe drinking water (%)* Habitat area suitable Water Annual flood duration (month) for protective species (km2)* Balance Extreme flood extent (km2) (water from Water Demand Waterlogged area (km2) river to GWL at end of dry season (m) ) Poor households affected Flood damage (Taka) Flood Agricultural by droughts, floods and salinity (%)* Damage production Displaced people due to disasters (%)* Rice production (Ton) Food security for the poor (%) Todays focus Food security Area affected by salinity (km2)* Cost of project implementation (Taka)

* Under development 5 Reflection on Previous Expert Session (29-04-2020)

More Information on Updating Technical Calibration & Validation Include river base flow Water Exchange Reference Whole Network (105 in MM between Land & River Documentation nodes) Final Waiting Final Ongoing

Separate Calibration & Update BDP2100 Development USER Validation scenarios with Reference Manual Dry & Monsoon Season newest insights (SIBDP) Ongoing (105 nodes) Final Final

Calibrate & Add information on Explore MM Validate Network BDP scenarios in MM Connection with Sector module in combination documentation Action Plans with WBM Final Final Ongoing Waterbalance

“To simulate waterdepths and shortages up to field level for each landtype per upazila in Bangladesh”

Simulation of: - Rainfall flooding: extents, depth, timing and duration - River and tidal flooding: extents, depth, timing and duration - Agricultural drought water availability / crop demand ratio - Hydrological drought impact on surface- and ground waterlevels

Closely connected with River network module: “To emulate transport of water through the major river network of Bangladesh” Water distribution Spatial and temporal information

• 10880 calculation units • (544) • Inside/outside BWDB project boundaries (2) • Agricultural land type map (BARC, 1999) + shrimp, forest, settlements, rivers and waterbodies (10) • Variables • Dominant soil category per upazila calculation unit (topsoil texture map, BARC) • Average land height (mPWD) from Bangladesh DEM (WARPO) • Average height of embankments per upazila is assumed to be 50cm higher than highest wl (1985-2017) from nearest river • Initial average groundwater levels derived from BADC • Rootzone storage is informed by effective soil depth map (BARC) • Percolation rates are informed by soil drainage map (BARC) • Precipitation (and reference evaporation) is based on timeseries of nearest BMD-station (1985 – 2017) 3 main steps in a decadal time step

Precipitation 1) Vertical distribution of water; 2) Horizontal Distribution of Water 3) Shortage of Excess of Evapotranspiration Water at Field Level

Drainage Inlet GW Irrigation

Infiltration + Percolation Module output for NW-region (base run)

Observations:

1. Clear seasonal diff II. Large annual variation III. Residual moisture supplemented by GW irrigation in beginning of dry season

Available for every district in Bangladesh (currently calibrated for NW-region) Module output for NW-region (base run)

Observations:

1. Clear seasonal diff II. Large annual variation III. Residual moisture supplemented by GW irrigation in beginning of dry season

Available for every district in Bangladesh (currently calibrated for NW-region) Example output: flooding and drought characteristics

Main observations:

I. Large annual variation II. Distinction in river floods and rainfall floods III. Timing and duration of floods inside/outside embankments IV. Crop water supply- demand ratio ~100%

Available for every district in Bangladesh (currently calibrated for NW-region) Extent of Network Module Flow Distribution

m1 Q_downstream-1 = B1 * (Q_upstream - QQo) Q_downstream-2 = Q_upstream - Q_downstream-1

Example case at Old Brahmaputra offtake:

1. QN80 = 0 when QN75 < 4900

m1 2. QN80 = B1*(QN75 - QQo1) when QN75 =< 21100

m2 3. QN80 = B2* (QN75 - QQo2) when QN75 > 21100

QN90 = QN75 – QN80 Water Level Calculation

Relation between Discharge and Water Level at Mawa, (N300) 8

7

6

5

When Q <= 42277 4

d1 a1 n1 h01 3

0 3000 1.7 0.1 WaterLevel (mPWD) 2

1 When Q > 42277 0 d2 a2 n2 h02 0 10000 20000 30000 40000 50000 60000 70000 80000 90000 100000 0 2500 1.9 0.3 Discharge (m^3/s) Jamuna River, Daudkandi Calibration and Validation

Node NSE PBAIS R_square .peakError Log_NSE Season​ Categoty N90 0.98 -0.48 0.98 3228.58 0.98 Monsoon​ Calibration​ N90 0.98 0.59 0.99 -755.86 0.93 Dry​ Calibration​ N90 0.94 -0.03 0.94 -873.84 0.95 Monsoon​ Validation​ N90 0.96 1.27 0.97 -440.57 0.93 Dry​ Validation​ River, Sujanagar

Calibration and Validation

Node NSE PBAIS R_square .peakError Log_NSE Season Categoty N280 0.99 1.83 0.99 -1443.89 0.98 Monsoon Calibration N280 0.94 10.82 0.98 1364.86 0.93 Dry Calibration N280 0.99 -0.82 0.99 -2424.77 0.99 Monsoon Validation N280 0.96 10.79 1.00 1161.84 0.97 Dry Validation , Ahmmadpur Calibration and Validation

Node MAE R_square Season Categoty N220 0.55 0.87 Monsoon Calibration N220 0.41 0.71 Dry Calibration N220 0.74 0.86 Monsoon Validation N220 0.31 0.77 Dry Validation , Gomnati Bazar Calibration and Validation

Node MAE R_square .peakError Season Categoty N185 0.44 0.75 0.14 Monsoon Calibration N185 0.43 0.13 0.68 Dry Calibration N185 0.46 0.78 0.36 Monsoon Validation N185 0.29 0.42 0.70 Dry Validation , Singimari Calibration and Validation

Node NSE PBAIS R_square .peakError LOG_NSE Season Categoty N62 0.41 4.08 0.43 2626.85 0.44 Monsoon Calibration N62 0.43 15.67 0.46 478.90 0.43 Dry Calibration N62 0.98 3.43 0.98 162.62 0.97 Monsoon Validation N62 0.94 7.41 0.95 40.46 0.55 Dry Validation , Proshadpur Calibration and Validation

Node NSE PBAIS R_square .peakError Log_NSE Season Categoty N176 0.25 -11.41 0.40 509.36 0.43Monsoon Calibration N176 0.22 9.91 0.30 227.27 -0.20Dry Calibration N176 0.20 -43.69 0.41 -84.35 0.14Monsoon Validation N276 0.62 33.10 0.81 198.31 0.09Dry Validation Verification of waterbalance results (NW-region)

Little data available for calibration: suggestions?

Source:

Bangladesh Integrated Water resources Assessment, 2014 Some flood characteristics () Some drought characteristics (Rajshahi division)

Source:

Bangladesh Integrated Water resources Assessment, 2014 Climate change scenarios (∆ method) Impacts BDP2100 climate change scenarios for NW-region

Scenario Base Productive Resilient Moderate Active Main observations: Climate situation 2020 2030 2050 2030 2050 2030 2050 2030 2050 Climate characteristics Sea level rise cm n/a 10 - 20 20 - 30 15 - 30 40 - 60 10 - 20 20 - 30 15 - 30 40 - 60 1. Much wetter in Temperature rise* ˚C n/a 0.5 1 1.5 2 0.5 1 1.5 2 Monsoon rainfall change % n/a 0 10 15 20 0 10 15 20 monsoon Dry season rainfall change % n/a 0 0 -10 -10 0 0 -10 -10 2. Larger & longer annual Peak discharge change % n/a 5 - 15 10 -20 15 - 30 20 - 40 5 - 15 10 - 20 15 - 30 20 - 40 Low flow discharge change % n/a -5 -15 -15 -30 -10 -25 -20 -40 floods, more extreme Decision support indicators Annual flood damage (infrastructure, housing, etc.) crore BDT 6.4 15% 37% 59% 88% 14% 37% 59% 88% floods Poor households affected by droughts and floods 522 7% 37% 59% 81% 7% 37% 59% 81% 3. Positive for GWL Annual rice crop production (T aman, Boro, Aus) Mtonnes 10.4 -1% -3% -5% -7% 0% -3% -5% -7% Food security for the poor: dietary energy supply calories/day n/a -22% -28% -25% -31% -22% -28% -25% -31% 4. Dry season low river Water system state indicators Total flood extent % of area 29% 8% 13% 21% 28% 8% 12% 21% 28% flow in Atrai basin Extreme flood extent % of area 47% 17% 22% 53% 64% 17% 22% 53% 64% decrease Flood duration days 114 1% 9% 12% 15% 1% 9% 12% 15% Dry season river flow (Atrai Basin) m3/s 25 7% 0% 0% -10% 4% -7% -3% -17% GWL decline per year cm 9 -3% -16% -22% -30% -3% -16% -22% -29% * effect on reference evapotranspiration Impacts BDP2100 climate change scenarios in time and space Impacts of BDP2100 climate change scenarios

Main observations:

1. More rainfall -> more runoff 2. Slightly less GW irrigation + more percolation

To do:

Visualize river inflow/outflow Impacts of climate change on river discharge Impact of climate change on low flow river discharge Conclusions – Impacts of climate change on NW

• The area becomes much wetter in the wet season, a bit drier in the dry season • Leads to larger flood extents, flood depths with little bit longer duration • Mainly leads to potentially more flood damage to infra and buildings, slightly more to rice crops • Poor households will be increasingly affected and the dietary energy supply will go down • Groundwater levels will rise, GW extraction remain stable, dry season water levels go down • Drought damage to crops does not change substantially

• No change in signal in timing of floods, as these are not included in the scenarios using the delta method Issues in Atrai – Hurasagar Basin

Siltation of main rivers Atrai and Hurasagar

Impeding navigation & erosion of river banks, increased flooding

FCD-projects (1980s)

Decreasing agriculture yield Decreasing fisheries Loss of floodplains & Source: Mathematical modelling for IWRM Chalan incl. Beel Halti (IWM, 2005) Cuts/breaches embankments BDP2100 note: increased flooding may also be caused by impediments to drainage due to unplanned infrastructure Closure of navigation routes Suggested SIBDP projects in Atrai – Hurasagar Basin

• Revitalization & restoration of Hurasagar and Atrai rivers (DP 1.3) • Dredging the Hurasagar and Atrai Rivers to increase discharge capacity and navigability • 30 km river bank protection works along Atrai River to protect growth-centers against erosion

• Revitalization and restoration of Beel Halti / (DP 1.2) • Option 5a) main public cuts with structures & closing all breaches, excavation Sib river, large structures -> low height embankments, + b) with additional FCD for Beel Halti area

• Implementation of rationalized water related interventions in Hurasagar basin (DP 1.21) • Program with measures: Infrastructure, knowledge, institutional Interventions modelled

Dredging (DP 1.3) Drainage capacity (DP 1.2) Combined (DP 1.3 + 1.2)

-1.2 m

-5 m Impacts of projects in Atrai – Hurasagar Basin

Main observations: Project Base DP 1.3a DP 1.3b DP 1.2a DP 1.2b DP1.3 + DP 1.2 Revitalization Atrai Revitalization Atrai Revitalization Beel Revitalization Beel 1. Dredging + improved Rajshahi division Without project and Hurasagar and Hurasagar Halti Halti Dredging - 1.20m + drainage improves Drainage capacity Drainage capacity - Drainage capacity + Option Dredging - 1.20m Dredging - 5.00m +25% 25% 25% flood situation for the BDP2100 scenario - Resilient 2020 2050 2020 2050 2020 2050 2020 2050 2020 2050 2020 2050 Decision support indicators short term Annual flood damage (infrastructure, housing, etc.) crore BDT 77 148% -8% 127% -9% 123% -7% 148% 27% 201% -15% 39% Poor households affected by droughts and floods 522 116% -14% 90% -17% 83% -21% 116% 63% 216% -36% -1% 2. Negative impact on Annual rice crop production (T aman, Boro, Aus) Mtonnes 89 -6% 2% -4% 3% -2% 1% -6% -1% -9% 3% 1% Food security for the poor: dietary energy supply calories/day n/a -27% 2% -25% 3% -23% 1% -27% -1% -29% 3% -21% GWL decline per River navigability* km/class n/a -- ++ - ++ - 0 ++ 0 + ++ + Rural access to safe drinking water* % n/a 0 0 0 0 0 0 0 0 0 0 0 year Habitat area suitable for protective species* km2 n/a + 0 + 0 + 0 0 0 0 0 + Displaced people due to disasters* % n/a -- 0 - 0 - 0 - 0 - 0 - 3. None of the Water system state indicators calculated measures Total flood extent % of area 21% 51% -18% 27% -33% 5% 1% 51% -1% 48% -18% 5% Extreme flood extent % of area 34% 124% -9% 64% -34% 37% 0% 124% -2% 123% -9% 31% is robust for extreme Flood duration days 85 25% -1% 26% -7% 25% -6% 25% 7% 27% -8% 10% Dry season river flow (Atrai Basin) m3/s 25 -5% 0% -5% 0% -5% 0% -5% 0% -5% 0% -3% climate change GWL decline per year cm 9 -32% 3% -29% 4% -26% 10% -32% -13% -48% 13% -3% Waterlogged area* km2 n/a ++ - ++ - 0 0 + 0 + 0 0 situation Environmental flow* m3/s n/a - 0 - 0 - 0 -- 0 -- 0 - Area affected by salinity* km2 n/a 0 0 0 0 0 0 0 0 0 0 0 Note: the role of Financial indicators Cost of project implementation mill BDT 0 112800 112800 78603 78603 152100 sediment is not included Cost of project maintenance mill BDT 0 32400 32400 0 0 32400 Total cost mill BDT 0 145200 145200 78603 78603 184500 * not yet quantified; expert opinion Salinity

• Salinity is one of the major problems in the South West part of Bangladesh. • The network module calculates surface water salinity intrusion, based on detailed MIKE21and MIKE 11 models from IWM.

• Calibration of salinity aspect is in progress. Salinity

For calculating salinity (in ppt) at each node, going downstream to upstream, following preliminary formula is developed:

푺풒ퟐ 푺풘ퟐ 푺풂풍풊풏풊풕풚풏풐풅풆 = 푺풒ퟏ풏풐풅풆 × 푸풏풐풅풆 × 푺풘ퟏ풏풐풅풆 × 푾풍 풏풐풅풆 × 푺풂풍풊풏풊풕풚풅풐풘풏풔풕풓풆풂풎풃풐풖풏풅풂풓풚

Where,

•sq1node = Discharge multiplication factor comes distribution factor •Qnode = Discharge at node point •Sq2node = Discharge exponential factor comes from distribution factor •Salinitydownstreamboundary = Comes from nearest downstream point •sw1node = Water level multiplication factor comes distribution factor •WLnode = water level at node point •sw2node = water level exponential factor comes from distribution factor Salinity: preliminary results Salinity: preliminary results Salinity

For calculating salinity (in ppt) at each node, going downstream to upstream, following preliminary formula is developed:

b) d S= (a * Q * (c * (WL- WLds) ) * Sds

Where, S = salinity a = multiplying factor for discharge b = exponential factor for discharge c = multiplying factor for water level d = exponential factor for water level

Sds = salinity of downstream node Impacts of salinity

1. Use a salinity impact function on agricultural production, fisheries (ESPA Deltas?) 2. Inlet of water in FCDI-project areas from river becomes dependent on salinity level in river; • if salinity river too high, regulators are closed -> no water inlet -> drought damage

• Do you have suggestions? PARTNERS

Follow the developments on: http://jcpbd.nl/index.php