f. fussi (1), l. fumagalli (1), t. bonomi (1), f. fava (1), b. di mauro (1), c.h. kane (2), m. faye...

F. Fussi(1), L. Fumagalli(1), T. Bonomi(1), F. Fava(1), B. Di Mauro(1), C.H. Kane(2), M. Faye(3), S. Wade(3), G. Faye(3), B. Hamidou (4) , R. Colombo(1)

(1) University Milano Bicocca, (2) University of Thies, (3)University Cheik Anta Diop – Dakar, (4) SNAPE - Conakry

contact: [email protected]

Use of remote sensing and terrain modeling to identify suitable zones for manual drilling in Africa and support low cost water supply

Groundwater, poverty and development, Overseas Development Institute, London 28/11/2014

STRUCTURE OF THE PRESENTATION

Introduction to manual drillingPrevious studies on suitable zones for manual drillingThe research projectGeneral research approachHydrogeological data processingRemote SensingResults achieved and next steps


Introduction to manual drilling

MANUAL DRILLING techniques of drilling boreholes for groundwater exploitation using

human or animal power (not mechanized equipment).

These techniques are well known in countries with large alluvial deposits (India,

Nepal, Bangladesh, etc)

High quality hand drilled wells can provide sustainable and

clean water supply


Introduction to Manual Drilling

Advantages and limitations

Advantages of manual drilling

Cheaper than mechanized boreholesEasy to implement with locally made equipment“manual work intensive” and not “capital intensive” ; source of income for local groups

Limitations of manual drilling

Manual drilling is feasible only under hydrogeological suitable conditionsSoft unconsolidated shallow geological layers Water level not too deep Good hydraulic conductivity of shallow porous aquifers

IT IS IMPORTANT TO IDENTIFY THOSE ZONES WHERE HYDROGEOLOGICAL CONDITIONS ARE SUITABLE FOR

MANUAL DRILLING

Introduction to Manual Drilling


Previous maps of suitablezones for manual drilling

BENIN BURUNDICENTRAL AFRICAN REPUBLICCHADGUINEA IVORY COAST LIBERIA MADAGASCAR MALI MAURITANIA NIGER SENEGAL SIERRA LEONE TOGO ZAMBIA

Since 2008 UNICEF has produced maps of suitable zones for manual drilling in 15 countries in Africa:

Previous studies on suitable zones


http://www.unicef.org/wash/index_54332.html

http://www.unicef.org/wash/index_54332.html

Previous method to identify suitable conditions for manual drilling (UNICEF)

Previous studies on suitable zones

Data sources:EXISTING MAPSNATIONAL DATABASE OF WATER POINTSQUALITATIVE EXPERIENCE OF WATER EXPERTS IN THE COUNTRY

Limitations:Input data are limited and often at very broad scale,inadequate for reliable interpretation (where data are limited) and more detailed studyNot systematic and structured procedure for data analysis. Difficult to compare results from different countries


COMBINED ANALYSIS OF 3 PARAMETERS:

THE UPGRO RESEARCH PROJECT

University Milano Bicocca (Italy)

University Cheik Anta Diop (Senegal)

SNAPE (Guinea)

UNICEF (Guinea and Senegal)

PARTNERS

FUNDED BY:


Use of remote sensing and terrain modeling to identify suitable zones for manual drilling in Africa and support low cost water supply

Main goal of the research

OBJECTIVE OF THE RESEARCHIntegration of direct hydrogeological information from existing database with indirect parameters from Remote Sensing and terrain modeling to characterize shallow aquifers and identify suitable zones for manual drilling

DURATIONNovember 2013 – April 2015


The research project

Research study area

REGION OF KANKAN AND FARANAH (EAST GUINEA)

REGION OF LOUGA – KEBEMER (NORTH WEST SENEGAL


The research project

Scientific approach

Geo-Environmental indicators (Geology, Soil, Morphometry, Vegetation dynamics, Soil moisture,

Thermal Inertia)

Thematic maps, Remote Sensing (optical, radar), Digital terrain model

(Geo)Statistical model

Borehole logs interpretation, pump test, geophysics

Map of suitable zones for manual

drilling

Hydrogeological features at observation points


General research approach

Methodological approach

Geology, Sols, etc

Topography &Geomorp

hology

Vegetation Type &

Persistence

Thermal Inertia

Soil Moisture

Base maps

STRM/ASTER DEM

MODIS Veg. Ind.

MODIS LST &

ALBEDO

RADAR ASAR

P, K, T, suitability class(estimated)

Classification (e.g. CART)

Borelogs - TANGAFRIC

Training set (P, K, T, class)

Testing set (P, K, T, class)

P, K, T, suitability class(observed)

Accuracy assessment

(e.g. Q2, RMSE)

cal

val

Spatial maps



Definition of a 2 steps procedure toestimate suitability for manual drilling


Feasibility: It depends on-depth of hard rock- depth of water level- thickness of hard lateritic layers

Potential for exploitation-It is obtained from assigning class of potential based on hydraulic transmissivity of porous unconsolidated aquifer from 0 to 50 m (considered the maximum epth for manual drilling) Estimation potential


Development of a specific software andcodification of hydrogeological data

DATABASE OF WATER POINTS SENEGAL AND GUINEA

STANDARDIZATION OF DATABASE STRUCTURE

IDENTIFICATION OF MOST FREQUENT CATEGORIES

ASSIGNING STRATIGRAPHIC CODES


Hydrogeological data processing

Processing of stratigraphic data


PROCESSING OF STRATIGRAPHIC AND PIEZOMETRIC DATA


Bonomi, T., (2009): Database development and 3D modeling of textural variations in heterogeneous, unconsolidated aquifer media: Application to the Milan plain. Computers & Geosciences 35 (2009) 134–145

Processing of hydrogeological data have been based on previous experience in Italy (Bonomi, 2009) and adapted to the context of Senegal and Guinea

Processing of stratigraphic data


EXTRACTION OF:

Each 2 m intervals:TEXTUREHYDRAULIC CONDUCTIVITY

For the whole log:DEPTH OF HARD ROCK,DEPTH OF WATER TABLETHICKNESS OF HARD LATERITIC LAYERSHYDRAULIC CONDUCTIVITY (K) OF UNCONSOLIDATED LAYERTRANSMISSIVITY

Processing of textural data at intervals of 2 m (Tangram software)


Extraction of hydrogeologicalparameters at borehole logs position


Depth of water

Depth of hard rock

Thickness of laterite

Average K (in exploitable layers)


Estimating feasibility and potentialfor manual drilling at logs position



SENEGAL STUDY AREA

Multitemporal analysis of ATI

Thermal Inertia (TI) is related to the characteristics of the surface materials as well as surface water content. Apparent Thermal Inertia (ATI) [K-1]can be estimated using solely remote sensing optical and thermal data (Van Doning et al., 2012 IJAEG).

- α0 is the albedo estimated from MODIS satellite optical data (MCD43B3 – 16 day – 1km)

- A is the amplitude of the diurnal cycle of the land surface temperature estimated from day/night MODIS thermal observations (MOD11C1/MYD11C1 – daily (4 overpass/day) – 1km).

- C is a solar correction factor.

Van Doninck J., Peters J., De Baets B., De Clercq E.M., Ducheynec E., Verhoesta N.E.C. (2011). The potential of multitemporal Aqua and Terra MODIS apparent thermal inertia as a soil moisture indicator. Int. J. App. Earth Obs. Geoinf. 13, 934–941

Remote Sensing


Multitemporal analysis of ATI

Daily time series of ATI (2012)

MODIS DATA (1km)MOD11C1/MYD11C1 +

MCD43B3

Data selection: no precipitation/clouds

Dry season ATI map

LOUGA

0.2 0.4 [K-1]

Dry ATI map 2012 (1km) - Northern Senegal

N

50 km

Groundwater, poverty and development,28/11/2014, Overseas Development Ini

Remote Sensing

Vegetation Dynamics The abundance, type and seasonal variations of vegetation are dependent from nutrient and water availability. Especially in drylands, vegetation can thus be related to water availability across the landscape.

Vegetation persistence or poor sensitivity from precipitation patterns may reflect the presence of shallow water and/or substrates suitable for deep plant roots development. Normalized Difference Vegetation Index (NDVI) time series were analysed to assess phenological indicators of vegetation cover and persistence during the dry season.

2007 2009 2010 2011 20122008YEAR

Remote Sensing


Vegetation Dynamics

MOD13Q1 (TERRA/AQUA MODIS - 16 day - Vegetation Indices, 250 m) 2008-2012

Filtering and smoothing procedure

Extraction of phenological metrics. Start of season, End of season, maximum, minimun, seasonal mean, etc

Vegetation persistence indicators (5 years average)- Mean dry season NDVI- Rate of NDVI decreas- Dry season lenght

LOUGA

DRY SEASON LENGHT (Number of days)

200-220220-240240-260260-280280-300

LOUGA

DRY SEASON MEAN NDVI

0.1 0.25

Remote Sensing


SOIL MOISTURE DYNAMICSSoil moisture dynamics may indicate areas with high water holding capacity or water accumulation.

ENVISAT-ASAR radar data (Global mode -1km) between 2008 and 2012 in C-band are used to estimate soil moisture by inversion of the radiative transfer model proposed by Karam et al., 1992.

Biomass maps derived from SPOT vegetation were used to separate the contribution of vegetation and soil moisture on the radar signal.

Karam M.A., Fung A.K., Lang R.H. and Chauhan N.S., 1992 : Microwave scattering model for layered vegetation, IEEE Trans. Geosci. Remote Sens., 30:767-784.

LOUGA

SOIL MOISTURE (%)

0 100

June 2010

Remote Sensing


Results achieved and next stepsMAIN RESULTS ACHIEVEDElaboration of a specific software to organize and process hydrogeological data based on the inventory of Senegal and GuineaDefinition of a structured method to estimate suitability for manual drillingProduction of maps of hydrogeological parameters and suitability for manual drilling at borehole logs positions in SenegalDefinition of procedures and generation of maps of soil moisture, phenology and ATI from optical and radar dataFirst set of direct field data (geophysics, pump test) collected and intepreted in SenegalNEXT STEPSCompleting field data collection in Senegal. And (depending from Ebola outbreak) in Guinea.Validation of hydrogeological interpretation with field informationDefinition of geostatistical model and integrated analysis of dataSpatialisation of hydrogeological interpretation from borehole logs position to the whole areaGenerating maps of suitability for manual drilling



Thanks very much for the attention

f. fussi (1), l. fumagalli (1), t. bonomi (1), f. fava (1), b. di mauro (1), c.h. kane (2), m. faye...

Documents