Robin Lynch Hydrology April 29,2010. Tlamacazapa, Guerrero Remote mountain village Extreme poverty Water scarcity Reliant on wells contaminated with arsenic.

Drought Frequency Analysis in Tlamacazapa, Guerrero, Mexico

Robin LynchHydrology

April 29,2010

Tlamacazapa, Guerrero

Remote mountain village

Extreme poverty

Water scarcityReliant on

wells contaminated with arsenic

Source: Google Maps

Previous Work

Concluded:Year round

piped water possible…but

Needs 10 meter high dam downstream

Source: Google Earth

Assumptions:Only used average rainfall between

2003-2007 Doesn’t take into account long term

variability!

0

50

100

150

200

250

Avera

ge R

ain

fall

(m

m/m

onth

)

Drought Frequency AnalysisDetermine how impacted Tlamcazapa is by

droughtsUse SPI (Standardized Precipitation Index)

Assumes data is normally distributed

Source: (Hallack-Algeria 2005)

Instituto Mexicano de Technologia del Agua (IMTA)16 gaging stationsWithin 25 kilometer radius of TlamaRainfall data between 1940-1998Inconsistent years in each station

First station

Source:(McKee et al. 1993)

Second Station

Whoa!!!

Theissen PolygonsCannot take average of SPI per yearEach gage’s SPI thrown off due to

incomplete time spanMust use Theissen Polygons to

choose Nearest NeighborFor years when closest gage has no

data, use nearest neighbor “hot decking” imputation procedure

60+ years of Data!Data from 1940-2007 (1944,1946,1999-2002 missing)

Check for Normality

…It’s Kinda Normal

SPI Values for Tlama: 1940-2007

HEC-SSP Only looks at probability of

exceedence

Probability Distribution Function

Drought Magnitude

1940

1942

1945

1948

1950

1952

1954

1956

1958

1960

1962

1964

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2004

2006

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

Drought Magnitude

Dro

ught

Magnit

ude (

Sum

of

SP

I)

Source: (McKee et al. 1993)

Recurrence Probability

P(SPI<0) = 0.5P(SPI<-1)= 0.15P(SPI<-1.5)= 0.1

Drought Frequency of Tlama

Single Year 2 Cons. Years 3 Cons. YearsMild Drought 2 4 8

Moderate Drought 7 44 296Severe Drought 10 100 1000

Extreme Drought 20 400 8000

Return Period

Future WorkStatistical regression to see the

relationship between drought and arsenic levels

Make recommendations for reservoir planning using this drought frequency analysis