Evaluation of Groundwater Sustainability in Selected ...€¦ · NUS Presentation Title 2001 • Human activities such as pumping and irrigation need large amounts of groundwater

NUS Presentation Title 2001

Evaluation of Groundwater Sustainability in Selected Regions of China and Future Projection under Climate Change

Zhou YiminA0105556H

Supervisor: A/P Yeh Jen-Feng Pat

9th May 2016

NUS Hydrology & Water Resources Group

1


Outline

• Introduction• Objective• Methodology• Procedures and Results

• Determine study area• Data evaluation• Results

• Discussion


2


Introduction – Why Study Groundwater Sustainability

• Groundwater is a limited resource.


Introduction Objective Methodology Procedures and Results Discussion

Source: Adapted from Figure 2, Freshwater Series No. A-2, Water - Here, There and Everywhere.

3

0.77%


• Human activities such as pumping and irrigation need large amounts of groundwater supply.

1.5-billion people rely on groundwater for drinking water supply (Karamouz, Ahmadi & Akhbari, 2011).

In cities such as Beijing or Tianjin, groundwater supplies more than 70% of total water resources (Beijing Municipal Water Conservancy Bureau, 2002; Dong et al., 2013).



Introduction – Why Study Groundwater Sustainability

4


Introduction – Why Study Groundwater Sustainability• Current Available Estimates of Global Groundwater Storage

and Groundwater Flux in Literature are of High Uncertainty.



Source Total groundwater storage (106 km3) Mean groundwater flux (103

km3/year)

L’Vovich (1967, 1974) 60.0 12.0

Korzun (1974) 23.4 13.32

UNESCO (1978) 13.2

National Council on Scientific Research USA (1986) 15.3UNESCO (1990) 23.4

Döll et al. (2002) 14.0

FAO (2003) 11.284

Döll and Flörke (2005) 12.882

Döll and Fiedler (2008) 12.7

Wada et al. (2010) 15.2

AQUASTAT (2011) 11.968

Richey et al. (2015) 7.0 (lower limit) & 23.0 (higher limit)

7.6 times

230% difference

24% difference

5


Objectives

• Establish groundwater resource database in China over the last several decades

• Evaluate groundwater availability and usage• Assess renewable groundwater stress (RGS) in selected

basins in China• Make comparison of RGSs calculated using different

definitions and data sources• Make suggestions on groundwater management

How sustainable groundwater (GW) is ?


Introduction Objective Methodology Procedures and Results Discussion 6


Methodology – GRACE


• Gravity Recovery and Climate Experiment (GRACE)

• Measure gravity anomalies• Three research institutes (CSR,

GFZ, JPL) provides three different sets of data applying different methods, models and coefficients.

• Level 3 product: global gridded water storage anomalies (in equivalent water height)



Methodology – GLDAS


• Global Land Data Assimilation System (GLDAS)

• Four global-scale hydrological models in GLDAS : Common Land Model (CLM), Mosaic (MOS), NOAH and Variable Infiltration Capacity (VIC).

• The model simulation output data from all four hydrologic models are used in this study.



Methodology – Moving Average Filter


• A 13-month moving average filter is applied.

• Average of GW anomalies from January 2005 to January 2006 in equivalent water height is calculated as the first moving average for July 2005.

• A series of moving average from July 2005 to January 2015 can be calculated by keeping moving the 13-month subsets forward.

• Eliminate short-term fluctuations and highlight long-term trend.



Renewable GW Stress (RGS)


𝑅𝐺𝑆 = %&()*%&+,+-.+/-.-01

(Richey et al., 2015)

GW use is defined as GW depletion.

GW availability is defined as mean groundwater recharge.



Renewable GW Stress (RGS) – GW Depletion

Natural (N), Anthropogenic (A) Surface Water (SW)Snow Water Equivalent (SWE)Canopy Water Storage (CWS)Soil Moisture (SM)Groundwater (GW)



Δ𝑆345 = Δ 𝑆𝑊+ 𝑆𝑊𝐸+ 𝐶𝑊𝑆 + 𝑆𝑀 + 𝐺𝑊 345

ΔGW = Δ𝑆345 − Δ 𝑆𝑊 + SWE + 𝐶𝑊𝑆 + 𝑆𝑀 345

11

𝑅𝐺𝑆 =𝐺𝑊𝑢𝑠𝑒

𝐺𝑊𝑎𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦

GRACE GLDAS


Renewable GW Stress (RGS) – GW Availability



Availability: mean groundwater rechargeModels: Community Land Model version 4.0

PCR-GLOBWB global hydrological model




Study Area

Source: UNESCO (2008)


• Severe groundwater quantity issues exist in China.



Study Area


• Severe groundwater quantity issues exist in China.

Country GW recharge(km3/year)

Population (million) (The World Bank, 2011) GW recharge (m3/year per capita)

Russia 788 142.96 5512

Canada 370 34.34 10775

China 829 1344.13 617

The USA 1384 311.72 4440

Brazil 1874 200.52 9346

Australia 72 22.34 3223

India 432 1247.45 346

Mean annual GW recharge per capita for seven largest countries in land area (2011).



Study Area



Qinling Mountain-Huaihe River Line

Yangtze River

550 million

100 million

650 million

Two geographical dividing lines between Northern and Southern China

Population (million)

Northern China ~600

Southern China ~700

15


Study Area



0100200300400500600700

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

GW

rec

harg

e (k

m3/

yr)

Year

Comparison of GW recharge in Northern and Southern China

North South

0

20

40

60

80

100

120

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

GW

use

(km

3/yr

)Year

Comparison of GW use in Northern and Southern China

North South

GW quantity issue in Northern China is much more serious than that in Southern China.

16


Study Area

North China Aquifer SystemSong-Liao BasinTarim Basin


Data availability

GRACEresolution (~200,000

km3)

More seriousproblem in Northern

China

Three study aquifers chosen



Study Area – North China Plain Aquifer (NCP)



• 34°N to 40°N, 113°E to 119°E

• Land area of 140,000 km2

• NCP covers Beijing, Tianjin, Hebei, Henan and Shandong provinces.

• Semi-humid Area

18


Study Area – Songliao Basin



• 40°N to 49°N, 120°E to 129°E


• NCP covers Heilongjiang, Jilin, Liaoning and Inner Mongolia provinces.

• Semi-humid Area

19


Study Area – Tarim Basin



• 36°N to 42°N, 75°E to 93°E


• NCP covers part of Xinjiang provinces.

• Arid Area with mean annual precipitation less than 100mm.

20


Data Evaluation – GRACE



Three GRACE observations are internally consistent with each other.

21





The peaks of TWS is delayed compared with the peaks of precipitation. The delayed time is from half an month to one month.

22





• Large difference between these two estimations• Sources of error for GLDAS: no component of GW, high

dependency on the accuracy of data inputs

Comparison of TWS by GRACE and GLDAS, take NCP as example

23


Data Evaluation

• Use GLDAS simulations for SM, SWE and CWS.• Similar to GRACE observations on TWS, four GLDAS

models (CLM, MOS, NOAH, VIC) have consistent results for SM, SWE and CWS.

• Calculate the average of three observations on TWS and average of four simulations for SM, SWE and CWS. The two averages are going to be used for further calculation.




Data Evaluation – Recharge



• The peaks for NCP and Songliao Basin appear in July and August each year (consistent with rainfall data)

• Tarim Basin has relatively stable but small amount of recharge as it is within arid zone.

PCR-GLOBWB modelsimulation on recharge in NCP, Songliao Basin and Tarim Basin

25


Results – GW Depletion



ΔGW = Δ𝑆345 − Δ SWE + 𝐶𝑊𝑆 345





ΔGW = Δ𝑆345 − Δ SWE + 𝐶𝑊𝑆 345

Aquifer name Depletion rate (mm/year)

NCP -8.83

Songliao Basin 7.13

Tarim Basin 1.16





ΔGW = Δ𝑆345 − Δ SWE + 𝐶𝑊𝑆 + 𝑆𝑀 345





ΔGW = Δ𝑆345 − Δ SWE + 𝐶𝑊𝑆 + 𝑆𝑀 345


NCP -8.86

Songliao Basin 3.67

Tarim Basin 0.972





ΔGW = Δ𝑆345 − Δ SWE + 𝐶𝑊𝑆 + 𝑆𝑀 345(With a 13-month moving average filter)





ΔGW = Δ𝑆345 − Δ SWE + 𝐶𝑊𝑆 + 𝑆𝑀 345(With a 13-month moving average filter)


NCP -7.19

Songliao Basin 3.15

Tarim Basin 0.636





ΔGW = u ∗ ΔH(u:specificyield,ΔH:changeinGWtable)From in-situ observations, only available in NCP.





Method Depletion in NCP(mm/year)

Depletion inSongliao Basin(mm/year)

Depletion inTarim Basin(mm/year)

Equation 2.3(include SM)

-8.83 7.13 1.16

Equation 2.4(exclude SM)(Depletion a)

-8.86 3.67 0.972

Moving averagefilter(Depletion b)

-7.19 3.15 0.636

In-situobservation ofGW table

-33.23 N.A. N.A.

Richey et al.(2015b)

-7.50 2.40 -0.23


Results – GW Recharge



Source NCP Songliao Basin Tarim BasinPCR-GLOBWBmodel (Wada, 2010)(Recharge a)

17.25 12.77 14.77

China WaterResources (2005)

111.16 69.24 37.81

Average of PCR-GLOBWB modeland China WaterResources bulletin(Recharge b)

64.21 41.01 26.29

Richey et al. (2015b) 96.56 20.16 -0.74

5.4 times difference


Results – Renewable Groundwater Stress



𝐷𝑒𝑝𝑙𝑒𝑡𝑖𝑜𝑛𝑎𝑅𝑒𝑐ℎ𝑎𝑟𝑔𝑒𝑎

𝐷𝑒𝑝𝑙𝑒𝑡𝑖𝑜𝑛𝑏𝑅𝑒𝑐ℎ𝑎𝑟𝑔𝑒𝑎

𝐷𝑒𝑝𝑙𝑒𝑡𝑖𝑜𝑛𝑎𝑅𝑒𝑐ℎ𝑎𝑟𝑔𝑒𝑏

𝐷𝑒𝑝𝑙𝑒𝑡𝑖𝑜𝑛𝑏𝑅𝑒𝑐ℎ𝑎𝑟𝑔𝑒𝑏

Richey et al.

(2015b)

NCP -0.514 -0.417 -0.138 -0.112 -0.08

Songliao

Basin

0.287 0.247 0.089 0.077 0.12

Tarim

Basin

0.066 0.043 0.037 0.024 0.32


Evaluation of Stress

Variable Stress Unstressed

Overstresshuman-

dominated stress

Positive Use(Gaining)

Negative Use(Extracting)

Positive availability(Recharging)



Negative availability(Discharging)

Stress Scale for VariableStress

0-0.1 Low

0.1-0.2 Moderate

0.2-0.4 High

Above 0.4 Extreme




Evaluation of Stress



𝐷𝑒𝑝𝑙𝑒𝑡𝑖𝑜𝑛𝑎𝑅𝑒𝑐ℎ𝑎𝑟𝑔𝑒

𝐷𝑒𝑝𝑙𝑒𝑡𝑖𝑜𝑛𝑏𝑅𝑒𝑐ℎ𝑎𝑟𝑔𝑒

𝐷𝑒𝑝𝑙𝑒𝑡𝑖𝑜𝑛𝑎𝑅𝑒𝑐ℎ𝑎𝑟𝑔𝑒𝑏

𝐷𝑒𝑝𝑙𝑒𝑡𝑖𝑜𝑛𝑏𝑅𝑒𝑐ℎ𝑎𝑟𝑔𝑒𝑏

Richey et al.

(2015b)

NCP Variable Stress

(Extreme)

Variable Stress

(Extreme)

Variable Stress

(Moderate)

Variable Stress

(Moderate)

Variable Stress

(Low)

Songliao

Basin

Unstressed Unstressed Unstressed Unstressed Unstressed

Tarim

Basin

Unstressed Unstressed Unstressed Unstressed Unstressed


Discussion – Difference on GW Depletion



Method Depletion in NCP(mm/year)

Depletion inSongliao Basin(mm/year)

Depletion inTarim Basin(mm/year)

Equation 2.3(include SM)

-8.83 7.13 1.16

Equation 2.4(exclude SM)(Depletion a)

-8.86 3.67 0.972

Moving averagefilter(Depletion b)

-7.19 3.15 0.636

In-situobservation ofGW table

-33.23 N.A. N.A.

Richey et al.(2015b)

-7.50 2.40 -0.23


Discussion – Difference on GW Depletion

Causes of differences:

• Over-estimation of in-situ GW anomalies due to the over-estimation of specific yield (𝛥𝐺𝑊 = 𝑢 ∗ 𝛥𝐻 ).

• Under-estimation of GW depletion due to model errors in GLDAS (no GW component).

• Other sources of GW: South-North Water Transfer Project (~5.7mm/year)




TGS

𝑇𝐺𝑆 =𝐺𝑊𝑆𝑡𝑜𝑟𝑎𝑔𝑒

𝐺𝑊𝑑𝑒𝑝𝑙𝑒𝑡𝑖𝑜𝑛𝑟𝑎𝑡𝑒

RGS𝑅𝐺𝑆 =

𝐺𝑊𝑢𝑠𝑒𝐺𝑊𝑎𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦

1RGS =

𝐺𝑊availability(𝑟𝑒𝑐ℎ𝑎𝑟𝑔𝑒)𝐺𝑊use(𝑑𝑒𝑝𝑙𝑒𝑡𝑖𝑜𝑛)

Comparison with Total Groundwater Stress (TGS)




Further Consideration

• Derive Renewable Groundwater Stress in time series instead of a time-averaged value

• Revise recharge models

• Integrate more climate change data

• Give recommendations on the limit of pumping based on groundwater stress




Sources: World Resources Institute

Groundwater stress Atlas


42

43

Documents

Evaluation of Groundwater Sustainability in Selected ...€¦ · NUS Presentation Title 2001 • Human activities such as pumping and irrigation need large amounts of groundwater