Content uploaded by Luis Ernesto Cervera-Gómez
Author content
All content in this area was uploaded by Luis Ernesto Cervera-Gómez on Jun 06, 2017
Content may be subject to copyright.
Nóesis. Revista de Ciencias Sociales y
Humanidades
ISSN: 0188-9834
noesis@uacj.mx
Instituto de Ciencias Sociales y Administración
México
Cervera Gómez, Luis E.
Impacts of Droughts on the Rio Conchos Watershed over the International Water Treaty Between
Mexico and the United States of America
Nóesis. Revista de Ciencias Sociales y Humanidades, vol. 17, núm. 33, enero-junio, 2008, pp. 116-
138
Instituto de Ciencias Sociales y Administración
Ciudad Juárez, México
Available in: http://www.redalyc.org/articulo.oa?id=85912926006
How to cite
Complete issue
More information about this article
Journal's homepage in redalyc.org
Scientific Information System
Network of Scientific Journals from Latin America, the Caribbean, Spain and Portugal
Non-profit academic project, developed under the open access initiative
Impacto de las sequías en el Río Conchos y sus repercusiones en el manejo internacional del agua entre México y Estados Unidos
Se analiza la sequía en el Río Conchos mediante el índice de sequía de
Palmer y datos de escurrimiento; en dos periodos decenales: 1950’s y
1990’s. Se relacionan con el Tratado Internacional de Aguas de 1944 en
dos de sus términos: condiciones de extrema sequía y períodos de cinco
años – preestablecidos como el tiempo máximo con condiciones de
sequía extrema.
is paper is aimed to analyze drought conditions using PDSI values
and hydrological data for the Rio Conchos Watershed during two decadal
periods: 1950´s and 1990s. Related to drought and from the International
Water Treaty signed in 1944 two keywords are analyzed: extreme drought
conditions and fi ve years cycle - pre-established as a maximum time with these
extreme drought conditions-.
Key
Words:
PALABRAS
CLAVE:
Rio Conchos Watershed,
drought and International
Water Treaty of 1944.
Descurrimiento del Río
Conchos, sequía y tratado
internacional de agua de
1944.
Luis E. Cervera Gómez1
1 Investigador de El Colegio de la Frontera Norte (mexicano), Dirección
Regional Noroeste. Candidato a doctor por la Universidad de Arizona.
Tels. (656) 617-65-93 y 617-89-58. Fax: (656) 618-40-55.
Correo: lcervera@colef.mx.
Impacto de las sequías en el Río Conchos y sus repercusiones en el manejo internacional del agua entre México y Estados Unidos
Impacts of Droughts on the
Rio Conchos Watershed over
the International Water Treaty
Between Mexico and the
United States of America
Fecha de recepción: 21 de abril de 2007
Fecha de aceptación: 07 de agosto de 2007
Nóesis
118
Introduction
The Conchos river basin, which drainage area entirely lies in Mexican
territory is the most important tributary of the Rio Grande/Rio Bravo
downstream of Fort Quitman, Texas. Since 1944 its flows together
with other five tributaries are part of a binational treaty on water al-
location between the United States and Mexico. Due principally to a
prolonged drought Mexico had failed to deliver the water granted to
the United States according to the Article 4 of the Water Treaty. As a
consequence Mexico has an accumulated deficit that looks hard to pay.
As part of this treaty, the concept of “extraordinary drought” is consid-
ered, but not measured. Thus, there is not a specific drought index to
define different levels of drought including extreme conditions. Other
relevant aspect of this treaty is the maximum time established as a
maximum time with drought conditions, a five-year cycle. To under-
stand the current international problems originated for drought, this
work analyze hydrological conditions throughout a statistical analysis
of the flows of the Conchos at the entrance of the two main dams (La
Boquilla and P. Madero) to test the following:
1. Changes in the mean and variability of the inflows during his-
torical records.
2. Analysis of indicators of drought, i.e. Palmer Drought Severity
Index (PDSI) for the closest climatic region in Texas (Near Oji-
naga).
3. Comparison of the 1950’s drought and the 1990’s drought. Sim-
ilarities, differences?
The final purposes of this research in terms to clarify basic con-
cepts of the 1944 International Water Treaty are twofold. First, the
understanding of what can be considered as extreme drought condi-
tions. Second, the understanding of duration of drought over time
compared with the five-year periods politically pre-established as a
maximum period of drought.
Impacts of Droughts on the Rio Conchos Watershed over the International
Water Treaty Between Mexico and the United States of America
119
Vol. 17 • número 33 • enero-junio 2008
Rio Conchos
The study area is represented by the Rio Conchos Watershed (see
Figure 1). This river is not only the Chihuahua’s most important river,
but it is also recognized as one of the most important river systems
in all of northern Mexico (Kelly, 2001). Because its waters give life
to an extensive territory in Mexico and it is also one of the principal
tributaries of the Rio Grande, this watershed has a crucial importance
as an international resource shared between Mexico and the United
States of America.
Climate and watershed characteristics
Using the Köeppen climatic system modified by Enriqueta Garcia
for the Republica Mexicana, the Rio Conchos has four climatic divi-
sions: (1) Very arid (BW) in the Chihuahua desert area; (2) Arid;
(3) Semiarid (BS1); and (4) Sub-Humid (AC) in the forests areas of
the Sierra Madre Occidental (SMO) (CNA, 1997). Rainfall in the
Rio Conchos watershed averages 377 mm (14.8 inches), but it ranges
from 300 (11.8 inches) to 1 000 mm (39 inches) at the upper basin to
about 200 (8 inches) to 400 mm (16 inches) in the lower basin (CNA,
Luis E. Cervera Gómez
Mexican portion of the
Rio Bravo/Grande Basin.
Rio Conchos
drainage area.
Figure 1. Rio Conchos Watershed. Map used under
permission of Israel Velasco, IMTA.
Nóesis
120
1997; CONABIO, 2000). The annual average temperature range is
from 8-18oC in the upper basin to about 16-22oC in the lower basin
(CONABIO, 2000).
In Mexico, the Rio Conchos Watershed has nine major populat-
ed municipalities (See Table 1) with about 1 068 901 inhabitants, of
which 63% are concentrated in the municipality of Chihuahua (main-
ly in the city of Chihuahua). There are 273 658 housing units with an
average of about 3.91 inhabitants per house.
In this basin, rapid rates of population and economic growth have
led to the widespread conversion of natural ecosystems to farmland,
industrial areas, and more urbanized areas. Urbanization, agricultural
intensification, resource extraction, and water resources development
are examples of human-induced phenomena that have had significant
impact on the people, the economy and the natural resources of the
Rio Conchos basin. Thus, higher rates of deforestation in the upper
basin, intensive agriculture, and urbanization in the middle and lower
basin areas are the principal factors undermining the natural condi-
tions of this basin. Consequently, the natural hydrological conditions
of this basin have been strongly affected in terms of hydrologic char-
acteristics, water quantity and water quality.
Impacts of Droughts on the Rio Conchos Watershed over the International
Water Treaty Between Mexico and the United States of America
Municipality
# of inhabitants
# of housing units
Avg. inhab./
house
Ojinaga
Hidalgo del Parral
Jimenez
San Fco. De Conchos
Delicias
Chihuahua
Camargo
Meoqui
Saucillo
24,313
100,881
38,259
2,837
116,132
670,208
45,830
39,848
30,593
6 568
24 509
9 260
748
29 466
173 582
11 574
10 228
7 723
3.68
4.11
4.13
3.78
3.93
3.85
3.95
3.89
3.91
Table 1. Municipalities Inside the Rio Conchos Watershed, Chihuahua, Mexico.
Source: INEGI, 2000. XII Censo General de Población y Vivienda:
preliminary results.
121
Vol. 17 • número 33 • enero-junio 2008
The Rio Conchos basin in its entirety contains one-half the entire
Rio Grande drainage in Mexico. Historically, this river has supplied
water mainly for agricultural activities (agriculture and livestock).
Furthermore, adequate streamflow is necessary to support riparian
habitats, and finally to satisfy downstream demands by the U.S. and
Mexico according to international agreements (USDOI, 1998). In the
Rio Conchos watershed, upstream from the subarea (Rio Grande-Rio
Conchos to Amistad Reservoir), expanding agricultural, mining, and
timber harvesting activities as well as urban and industrial develop-
ment affect both the quantity and quality of Rio Grande flows and its
aquatic-biological characteristics within the subarea (USDOI, 1998;
Davis, 1980).
At the confluence of the Conchos with the Rio Grande (Presi-
dio/Ojinaga), the inflow of the Conchos increase significantly the
Rio Grande’s streamflow. Thus, the Rio Conchos supplies the largest
percentage of Rio Grande flows allocated by Mexico in accordance
with the international water treaty (USDOI, 1998). During the 1980’s
the total annual flow of the Rio Conchos averaged 737 000 acre-feet
(908.7 Mm3), representing this flow approximately five times the flow
of the Rio Grande measured upstream (IBWC, 1989 in USDOI,
1998).
e 1944 International Water Treaty
In 1944, the U.S and Mexico signed a treaty to allocate the water re-
sources from the international watersheds shared by these two coun-
tries. In general, this treaty includes the Colorado River, the Tijuana
River and the Rio Bravo/Rio Grande. The Rio Grande/Rio Bravo
waters between Fort Quitman, Texas and the Gulf of Mexico are
“hereby allotted to the two countries in several manners, but in this
research I will only focused to the waters granted to the United Sates
under Article 4 of the Water Treaty.
In accordance with the 1944 Water Treaty, the United States has
right to a portion of water coming from six tributaries. This granted
water is described in the Subparagraph (c) of Article 4 as follows:
Luis E. Cervera Gómez
Nóesis
122
One-third of the flow reaching the main channel of the Rio Grande from the
Conchos, San Diego, San Rodrigo, Escondido and Salado Rivers and the Las
Vacas Arroyo, provided that this third shall not be less, as an average amount in
cycles of five consecutive years, than 350,000 acre-feet annually (IBWC, 2002).
However, same Article 4 considers in its last part, what to do in
case that Mexico fail to pay the aforesaid water allocation. Thus, the
Water Treaty literally states:
In the event of extraordinary drought or serious accident to the hydraulic
systems on the measured Mexican tributaries, making it difficult for Mexico to
make available the run-off of 350,000 acre-feet (431,271,000 m3) annually,
allotted in subparagraph (c) of paragraph B as the minimum contribution from
he aforesaid tributaries, any deficiencies existing at the end of the aforesaid
five-year cycle shall be made up in the following five-year cycle from the said
measured tributaries.
Persistence of drought in Chihuahua for a ten years period from
1990 to 2000 had seriously impacted water quantity generated from
the Rio Conchos and as a consequence less water was reaching the
Rio Bravo. This situation put a Mexico in a deficit situation, which
affects the 1944 U.S./Mexico water treaty. Kelly (2001) reports that in
the five-years cycle ending on October 2, 1997, Mexico owed about
1,240 Mm3 (1.024 million acre-feet). In the current five year cycle
corresponding to the period from October 3, 1997 to April 6, 2002,
Mexico has an accumulated deficit of 1 476 181 acre-feet (1,820.13
Mm3) (IBWC, 2002).
Drought
1) Changes in the mean and variability of the inflows
during the period of record
For this part, I did a statistical analysis of the Rio Conchos flows at the
entrance of the two main dams; these are La Boquilla, and P. Madero.
A 63 years spanning period of inflows from 1935 to 1998 were con-
sidered from La Boquilla. Also, a 49 years spanning period from 1949
to 1998 were analyzed for the Madero reservoir. Statistical analysis
Impacts of Droughts on the Rio Conchos Watershed over the International
Water Treaty Between Mexico and the United States of America
123
Vol. 17 • número 33 • enero-junio 2008
comprises descriptive parameters to study patterns and changes in
flows regime. These stats are mean, coefficient of variation, maximum,
minimum, standard deviation, skew and Kurtosis. All descriptive data
were analyzed using EXCEL spreadsheets and results are displayed
in Appendixes I and II. Additionally to test statistical differences be-
tween drought conditions during 50’s and 90’s the SPSS was used.
The annual flows of the Rio Conchos before to reach La Boquilla
dam and the Madero reservoir are highly variable, this because the
most important climatic factor affecting Chihuahua’s rivers is the spa-
tial and temporal variable pattern of its rainfall. Figures 2a and 2b
show the inflows of the two studied reservoirs. As we can observe river
flows of the Rio Conchos varies greatly from year to year representing
a pattern of possible droughts, floods and what can be considered nor-
mal conditions. In La Boquilla, the mean annual flow was estimated
in 1 229 39 millions of cubic meters (Mm3). The maximum registered
annual flow was about 3 529 20 Mm3 (almost three times the average
flow) during 1991. The minimum annual flow was estimated in 137
Mm3, corresponding this figure to the year of 1951 when the region
was having extreme drought conditions. According to Fierro (1999)
during 1951 the average annual precipitation was about 129.8 mm for
the State of Chihuahua.
Luis E. Cervera Gómez
Figure 2a. The annual inflows to La Boquilla Reservoir, Rio
Conchos (1935-1998)
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
1935 1939 1943 1947 1951 1955 1959 1963 1967 1971 1975 1979 1983 1987 1991 1995
Year
Figure 2a. The annual inflows to La Boquilla Reservoir, Rio Conchos (1935-1998)
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
1935 1939 1943 1947 1951 1955 1959 1963 1967 1971 1975 1979 1983 1987
1991 1995
Year
Figure 2a. The annual inflows to La Boquilla Reservoir, Rio Conchos (1935-1998)
Nóesis
124
Considering the mean streamflow value of about 1229.39 Mm3 as
normal conditions of flow and from a hydrologic perspective we can
detect a kind of hydrological drought or extreme hydrological drought
conditions during 1940, 1948, 1950-1951, 1956-1957, 1959, 1969,
1982, 1985, 1994-1995, and 1997-1998. Also, we can emphasize wet
or extreme wet conditions during 1938, 1941-1942, 1958, 1966, 1968,
1978, 1981, 1984, 1986, and 1991 (three times the mean value). Fi-
nally, the annual flow variability presents a coefficient of variation of
about 60%.
In the Madero reservoir, the mean annual flow was estimated in
400.81 millions of cubic meters (Mm3). The maximum registered an-
nual flow of about 941.30 Mm3 (more than two times the average
flow) was presented during 1981. The minimum annual flow was es-
timated in 36.70 Mm3, corresponding this figure to the year of 1994
when the region was having extreme drought conditions (this is also
analyzed in the next task). The impacts of drought during 1994 were
very drastic for the region, in such a way that the 50% of the livestock
inventory (1 000 000 heads) was lost. Also, great problems were re-
ported for the temporal and irrigated agriculture.
Considering the mean value of about 400.81 Mm3 normal condi-
tions of flow we can detect drought or extreme hydrologic drought
conditions during 1950-1951, 1952-1954, 1956, 1961-1962, 1964-
1965, 1967-1969, 1977, 1982-1983, 1985, 1992-1995, 1997-1998.
In addition, we can emphasize wet or extreme wet conditions during
1949, 1955, 1958, 1960, 1966, 1968, 1971-1972, 1974, 1978, 1980,
Impacts of Droughts on the Rio Conchos Watershed over the International
Water Treaty Between Mexico and the United States of America
Figure 2b. The annual inflows to Madero Reservoir, Rio Conchos (1949-1998)
125
Vol. 17 • número 33 • enero-junio 2008
1981, 1984, 1986, 1990-1991, and 1996. Finally, the annual flow vari-
ability presents a coefficient of variation of about 63%.
In La Boquilla dam, the greater mean monthly inflows start in July
to end in October. From the summer season, the maximum values are
during August and September (352.91 and 376.01 Mm3 respectively).
In the Madero reservoir, we found that the maximum flow values are
during July to September, reaching a maximum during August (125.2
Mm3), see figure 3.
2) Analysis of indicators of drought, i.e. PDSI for the closest
climatic region in Texas
2.1 Definition of drought.— It is widely accepted that drought is a nor-
mal, recurrent feature of climate, and it can be present in most of the
different climatic zones. Furthermore, it is also generally accepted that
there is not a general definition of drought. Thus, we can find more
than 150 different concepts, which can be based on meteorological,
agricultural, hydrological, and socioeconomic disciplinary perspectives.
Basically we can find two main types of drought definitions, the first
one is conceptual and the second one is operational. The first kind of
definitions is formulated in general terms, and its utility relies in that
it helps people to understand the meaning of drought. According to
the National Drought Mitigation Center (NDMC), the operational
concepts are useful to detect the beginning, end, and degree of severity
Luis E. Cervera Gómez
0
50
100
150
200
250
300
350
400
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
La Boquilla (1935-1998) Madero (1949-1998)
Figure 3. Mean monthly flow at the Rio Conchos (1949-1998)
Nóesis
126
of a drought. Furthermore, the operational definition can be used to
analyze drought frequency, severity, and duration for a given period.
In this research to analyze drought conditions I am using the
Palmer Drought Severity Index (PDSI). The PDSI in an important
climatological tool to evaluate the scope, severity, and frequency of
prolonged periods of abnormally dry or wet weather (Climatic Predic-
tion Center, 2000). This PDSI was designed to characterize drought
solely in terms of meteorological phenomena. Thus, monthly precipi-
tation and monthly temperature are used to estimate it. According to
Steila (1972), Palmer defines a drought period as:
An interval of time, generally of the order of months or years in duration, during
which the actual moisture supply at a given place rather consistently falls short
of the climatically expected or climatically appropriate moisture supply. Further,
the severity of drought may be considered as being a function of both the duration
and magnitude of the moisture deficiency.
After this definition, it is necessary to analyze PDSI in terms of
both duration and magnitude of dry or wet conditions. For the PDSI,
11 categories of wet and dry conditions are defined (Table 2).
Impacts of Droughts on the Rio Conchos Watershed over the International
Water Treaty Between Mexico and the United States of America
Category
Dryness or Wetness conditions
4.0 and above
3.0 to 3.99
2.0 to 2.99
1.0 to 1.99
0.5 to .99
0.49 to –0.49
-0.50 to –0.99
-1.0 to –1.99
-2.0 to –2.99
-3.0 to –3.99
-4.0 and below
Extreme moist spell
Very moist spell
Unusual moist spell
Moist spell
Incipient moist spell
Near normal
Incipient drought
Mild drought
Moderate drought
Severe drought
Extreme drought
Table 2. PDSI Values for the 11 Drought (or wet) Categories
127
Vol. 17 • número 33 • enero-junio 2008
The mean monthly variation of PDSI estimated for the period
1895 to 2000 is presented in Figure 4. This chart may represent the
general drought conditions of the arid and semiarid areas of the Rio
Conchos basin over time. Considering the average PDSI values from
1895 to 2000, the Plamer’s Index places January to June and August
and October to December in the same drought category (0.49 to -
0.49); this category represent a “near normal” condition. The wetter
PDSI category occur during the summer months (June, July and Au-
gust), here it is observed an increase in moisture conditions. During
the rainy season the peak value is reached during July, which is con-
sidered as incipient moist spell. In general the near normal condition
of drought may represent the average drought conditions for the State
of Chihuahua, Mexico. Furthermore, PDSI recorded a maximum of
“incipient drought” intensity, which reached its peak in September.
To give us some ideas in practical terms about what the different
stage conditions of drought can represent, we can see what was re-
ported by Steila (1972) for the state of Arizona.
Luis E. Cervera Gómez
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0.0
0.1
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
Mean
Figure 4. Mean Palmer Indices of Drought (1895-2000): Texas 05
Nóesis
128
Magnitude and duration of dry and wet conditions—. The PDSI val-
ues are highly variable over time. As we already know, these drought
indexes are computed using precipitation and temperature; these are
highly variable climatic factors, very characteristic of arid and semiarid
regions. Figure 5, is showing the PDSI values for the 1895-2000 pe-
riod. PDSI values represent the magnitude of a dry or wet condition.
From this figure and just analyzing data that match with the hydro-
logic data (river flows) we found the following dry or wet conditions.
Impacts of Droughts on the Rio Conchos Watershed over the International
Water Treaty Between Mexico and the United States of America
Drought category
Description*
Mild drought
“In the early part of the year, ranges are fair
and soil moisture generally adequate. In late
spring and summer, stock tanks are drying,
ranges are dry, and fire hazards in grass and
forest grazing land becomes apparent.”
Moderate drought
“Ranges are very dry and beginning to
deteriorate, stock water is short, and rains are
needed. Fire hazards increase.”
Severe drought
“Ranges are extremely dry, water supply
scarce, and the hauling of stock water and
supplemental feed is common. The threat of
fire is very high.”
Extreme drought
“Parched vegetation, hauling of water and
extensive pumping needed, and range animals
are loosing weight and suffering death
losses.”
*These drought descriptions were based on 40-year period, for each
of the drought category.
Table 3. Descriptive References of Main Drought Categories for the State of Ari-
zona Reported by Steila (1972)
129
Vol. 17 • número 33 • enero-junio 2008
Having an extreme drought class we can identify the years of 1935,
1953, 1957, 1997, and 2000. The years of 1951-1954, 1965, 1971,
1975, 1991, 1995-1996 are considered with severe drought. With mild
to moderate drought we can detect the following years: 1937, 1938-
1940, 1945-1948, 1951, 1963-1964, 1973, 1978, 1985, 1990, 1993-
1995, and 1998. Conversely, with extreme wet conditions we can see
the years of 1941, 1975, 1979, 1981, 1987, and 1993. As very moist
spell conditions are considered the years of 1959, 1979, and 1981.
Maybe the most important aspect of a drought condition is its per-
sistency over time. In this analysis, we are considering the consecutive
months with dry conditions. Figure 6 is showing duration of drought
in months for the studied region. From this figure, it is easy to detect
years as 1915, 1927, 1933, 1950-1951, 1961, 1992, and 1997 that have
longer duration periods of drought.
Extreme drought.—
It is a very serious situation which results from many months, or even years,
of abnormally dry weather. During extreme drought, agricultural crops are
complete failure; industries and municipalities may face the need for rationing
water; and the local and regional economy begins to become disrupted. Thus
extreme drought is essentially a disaster.
Luis E. Cervera Gómez
-9
-6
-3
0
3
6
9
12
1895 1903 1911 1919 1927 1935 1943 1951 1959 1967 1975 1983 1991 1999
Year
Figure 5. Palmer Drought Severity Index (1895-2000): Texas 05
Nóesis
130
3) Compare the 1950’s drought and the 1990’s drought.
Similarities, differences?
To compare 1950’s versus 1990’s drought conditions, in this section I
am only considering data corresponding to these two decades. Figure
7 is showing the PDSI values for two periods, from 1950 to 1960
and from 1990 to 2000. First, data indicates that 1950’s drought was
more persistent than the 1990’s, but the severity index of both decades
was similar. In the year of 1950, initiate the longest period of drought
during a total of 86 months of consecutive dry months. In the year
of 1990, initiate another extensive period of drought having 45 con-
secutive dry months (Figure 8). From Figure 8, we found a similar
condition regarding to total duration of drought between the 1950’s
and 1990’s drought periods. Thus, between 1950 and 1950 we have
a total of 87 consecutive months and from the 1992 to 1998 period
we have a total of 89 consecutive months of dry conditions. The last
comparison between 1950’s and 1990’s drought conditions was made
using two histograms of frequency of drought for every decade period.
These histograms are represented by figures 9a and 9b. In general,
apparently there were more extreme drought and moderate drought
during the 1950’s. Conversely, there were more wet conditions dur-
ing the 1990’s, having this period very moist and extreme moist spell
conditions. Finally, considering the total inflow volume for one of the
reservoirs (La Boquilla), we found that the total volume received dur-
ing the 1950’s was about 9 617 5 Mm3 compared to the 11 602 1 Mm3
received in the 1990’s.
Impacts of Droughts on the Rio Conchos Watershed over the International
Water Treaty Between Mexico and the United States of America
0
20
40
60
80
100
1895 1903 1911 1919 1927 1935 1943 1951 1959 1967 1975 1983 1991 1999
Year
Figure 6. Duration of Drought (PDSI: 1895-2000): Texas 05
PDSI: Magnitude
131
Vol. 17 • número 33 • enero-junio 2008
As a concluding remark I found a visible common aspect between
the 1950’s and the 1990’s drought; both decadal periods have the most
extensive consecutive dry condition.
Luis E. Cervera Gómez
-6
-3
0
3
6
9
1950 1951
1952
1953 1954 1955 1956 1957 1958 1959 1960 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
PDSI (Magnitude)
YEAR
Figure 7. Palmer Drought Severity Index (1950’s and1990’s): Texas 05
86
11
21
45
4 4
36
0
20
40
60
80
100
1950 1951 1953 1955 1956 1958 1988 1990 1992 1993 1995 1997 1998 2000
Year
Duration (month)
YEAR
Figure 8. Duration of Drought (PDSI: 1950’s and1990’s): Texas 05
Year
M
onths
Year
M
onths
Year
M
onths
Year
M
onths
Year
M
onths
Years
M
onths
1895
2
1908
29
1926
33
1945
8
1965
2
1981
3
1896
7
1911
4
1929
15
1946
8
1966
10
1982
4
1897
15
1912
14
1933
44
1947
19
1968
9
1983
7
1898
10
1913
2
1936
5
1949
7
1969
6
1985
10
1899
7
1915
37
1937
4
1950
86
1970
12
1986
21
1900
25
1920
8
1938
7
1956
11
1973
7
1990
45
1903
13
1921
7
1939
7
1957
3
1974
19
1994
8
1905
3
1922
7
1942
9
1958
57
1977
11
1995
36
1906
2
1923
3
1943
3
1963
2
1979
12
1996
4
1907
16
1924
14
1944
4
1964
13
1980
12
1997
4
1998
36
Table 4. Initiation of Consecutive Dry Months: Texas 05, Near Ojinaga
Nóesis
132
Analytical comparisons between decadal periods of drought
To analyze if there are statistical differences between drought condi-
tions during 1950’s versus 1990’s I did a simple comparison between
the means values. Thus, considering that in the Rio Conchos water-
shed the rainy seasons correspond to the months of July, August and
September, a mean three-monthly PDSI value was computed for ev-
ery decadal period. After that I did a comparison between means us-
ing the one tail T-student probability function.
0
5
10
15
20
25
Extreme
drought
Severe
drought
Moderate
drought
Mild
drought
Incipient
drought
Near
normal
Incipient
moist spell
Moist spell Unusual
moist spell
Very moist
spell
Extreme
moist spell
Class
Figure 9a. Frequency of Drought (1950-1960): Texas 05
0
5
10
15
20
25
Extreme
drought
Severe
drought
Moderate
drought Mild Incipient Near
normal
Incipient
moist spell
Moist spell Unusual
moist spell
Very moist
spell
Extreme
moist spell
Class
Figure 9b. Frequency of Drought (1900-2000): Texas 05
Impacts of Droughts on the Rio Conchos Watershed over the International
Water Treaty Between Mexico and the United States of America
0
5
10
15
20
25
Extreme
drought
Severe
drought
Moderate
drought
Mild
drought
Incipient
drought
Near
normal
Incipient
moist spell
Moist spell Unusual
moist spell
Very moist
spell
Extreme
moist spell
Class
0
5
10
15
20
25
Extreme
drought
Severe
drought
Moderate
drought
Mild
drought
Incipient
drought
Near
normal
Incipient
moist spell
Moist spell Unusual
moist spell
Very moist
spell
Extreme
moist spell
Class
133
Vol. 17 • número 33 • enero-junio 2008
Table 5 exhibits the mean PDSI values computed with three sum-
mer months for the two decadal period. The mean PDSI values for
the decadal period during fifties was computed in –1.824 versus the
–0.055 mean PDSI value registered during 90’s. Looking at the PDSI
scale, the first value corresponds to Mild Drought conditions and the
second corresponds to Near Normal conditions. Comparing these two
decades throughout a paired test we have the following conclusions.
Although the difference in means was equal to –1.769, we are unable
to report a difference between means for the fifties and nineties peri-
ods (t=-1.411, p value= 0.189).
Luis E. Cervera Gómez
Year July August September Mean PDSI
1950 1.37 -0.48 -0.1 0.26
1951 -2.08 -2.52 -3.07 -2.55
1952 -1.65 -2.81 -3.41 -2.62
1953 -4.1 -4.5 -5.1 -4.56
1954 -3.23 -2.17 -3.09 -2.83
1955 -2.07 -2.3 -2.53 -2.30
1956 -3.94 -4.3 -4.78 -4.34
1957 -3.5 -3.49 -3.94 -3.64
1958 1.91 1.54 2.71 2.05
1959 -0.64 -0.62 -1.39 -0.88
1960 1.66 1.78 0.63 1.35
1990 1.69 2.72 3.7 2.70
1991 3.12 3.18 4.19 3.49
1992 7.63 7.09 -0.64 4.69
1993 -0.55 -0.55 -0.96 -0.68
1994 -2.35 -3.11 -3.31 -2.92
1995 -2.86 -3.35 -2.75 -2.98
1996 0.02 1 1.21 0.74
1997 -0.03 -0.26 -0.88 -0.39
1998 -3.03 -2.48 -3.31 -2.94
1999 -1.1 -1.63 -2.19 -1.64
2000 -3.83 -4.38 -5.22 -4.47
Table 5. Summer PDSI Values: Texas 05, Near Ojinaga
Nóesis
134
In a second analysis a Two-sample test was computed and as a
result of the T test procedure we have two density plots as showed
in Figure 10. On the far left and right sides of the density plot for
each test variable. The middle portion of each graph shows the actual
distribution of data points, with a normal curve for comparison. The
standard deviation differs considerably (2.232 and 2.917). The box
plot on the left is for the 1950’s. In this case the peak of the curve are
no very close from the median values (horizontal line). This indicates
that this curve is asymmetrical. 1990’s PDSI values show a higher
standard deviation, but more symmetrical having higher variance than
1950’s data.
Testing decadal fl ow data against historical monthly averages
To know if the average monthly streamflow values are different from
the average monthly historic means registered in every reservoir (Bo-
quilla, Madero) and Ojinaga we made the following One-sample t test.
The mean monthly registered stream flow entering to La Boquilla
reservoir is about 102.45m3 considering the period from 1935 to 1998,
with a maximum of 1,539 and a minimum of 0; the mean monthly
stream flow registered in the streamflow entering into Madero dam
Impacts of Droughts on the Rio Conchos Watershed over the International
Water Treaty Between Mexico and the United States of America
1990’s
1950’s
Decade
-5
-4
-3
-2
-1
0
1
2
3
4
5
PDSI
0
1
2
3
4
5
6
7
8
Count
0
1
2
3
4
5
6
7
8
Count
Figure 10. PDSI density plots: Texas 05, near to Ojinaga
135
Vol. 17 • número 33 • enero-junio 2008
was about 33.40m3, with a maximum of 596 and a minimum of 0;
finally, the mean monthly registered stream flow registered in the Oji-
naga was about 70.38m3, with a maximum of 1 285 5 and a minimum
of 6.30m3.
Testing data from the Boquilla dam during 50 and 90
During 50’s we are able to say that the historical mean (102.45)
does differ significantly from the mean value for this decadal period
(72.857) (t = –1.67, p value= 0.126). Conversely during 90’s there is
not significant difference from the historical mean (102.45) against
the mean value for this decadal period (114.554) (t= 0.512, p value=
0.620).
Testing data from the Madero dam during 50 and 90
During 50’s we are able to say that the historical mean (33.40) does
not differ significantly from the mean value for this decadal period
(29.618) (t = –0.706, p value= 0.49). Same happen during 90’s since
there is not significant difference from the historical mean (33.41)
against the mean value for this decadal period (28.125) (t=-0.819, p
value= 0.432).
Testing data from the Ojinaga streamfl ow during 50 and 90
During 50’s data were limited, so for this decade we are using data
from 1955 to 1960. Assuming that this period represent the 50’s we
are able to say that the historical mean (70.38) does not differ sig-
nificantly from the mean value for this five years period (72.648) (t =
0.087, p value= 0.934). Although in this case we have complete data
same happen during 90’s since there is not significant difference from
the historical mean (70.38) against the mean value for this decadal
period (70.653) (t=0.012, p value= 0.990).
Luis E. Cervera Gómez
Nóesis
136
Concluding remarks
Climatic drought is very characteristic of the middle and lower areas of
the Rio Conchos Basin, which drainage area is inside the Chihuahuan
desert region except for the upper basin located on the Sierra Madre
Occidental. In this region, droughts had been historically reported
since 1576 affecting seriously to the population and its economics ac-
tivities —agriculture and livestock mainly. More recently severe and
prolonged droughts had been registered in the Rio Conchos Basin
especially during the 1930’s, 1950’s and 1990’s. Evidence suggests that
it is the last decade of drought that more had affected hydrological
conditions lowering its surface runoff, and water stocks of main dams
such as La Boquilla and Madero dams. As a consequence less water is
reaching the Rio Grande near Ojinaga, causing problems to Mexico to
provide unless the minimum flows allotted to the United States under
the 1944 International Water Treaty between these two countries.
Now, Mexico has a significant water debt, which looks hard to pay
if severe drought conditions persist.
In this work, drought was analyzed using two approaches. The first
one includes the historical analyses of stream runoff measured before
the entrances of the La Boquilla and Madero dams. The second analy-
sis was focused in a statistical analysis of the Palmer Severity Drought
Index (PDSI) measured in a Texas climatic station near to Ojinaga.
Analysis indicates that one of the main characteristics of the stream
runoff is its high variability, presenting the annual flow a coefficient
of variation of about 60% for surface runoff entering the Boquilla and
Madero dams. Thus, the high variability of the Rio Conchos flows
are evident presenting great changes year to year revealing us periods
with drought, floods and normal conditions. Special attention requires
hydrological droughts during 50’s and 90’s.
Analyzing PDSI data we have that the most frequent condition
of drought for Chihuahua near Ojinaga region is a state of incipient
drought, followed by a mild drought condition. Results indicate that
under extreme drought conditions we have the following years: 1935,
1953, 1957, 1997, and 2000. Severe drought conditions were present
Impacts of Droughts on the Rio Conchos Watershed over the International
Water Treaty Between Mexico and the United States of America
137
Vol. 17 • número 33 • enero-junio 2008
during 1951-1954, 1965, 1971, 1975, 1991, 1995-1996. The longer
periods of drought were detected the following years: 1915, 1927,
1933, 1950-1951, 1961, 1992, and 1997.
In this research it was found that 1950’s drought was more persis-
tent than the 90’s, although the severity index was similar. Thus, dur-
ing 1950, initiate the longest period of drought totalizing 86 (7 years)
dry months. Furthermore, 45 (≅ 4 years) consecutive dry months ini-
tiated in 1990. During the period from 1992 to 1998 we have a total
of 89 (≅ 7.4 years) months having dry conditions. Significant differ-
ences were found between the 1950 and 1990’s, having more extreme
to moderate drought conditions during 50’s. There is evidence that
there are longer drought periods having more than five years with
severe to extreme conditions. This is important to note because under
these conditions it could be more difficult for Mexico to do one’s duty
regarding water agreements.
Testing data from the Boquilla dam, statistical analysis indicates
that the average stream runoff values during 50’s differ significantly
from the historical mean. Conversely, during 90’s there is not signifi-
cant difference from the historical mean. For the Madero dam there
were no significant differences between both decadal periods (50’s and
90’s) against the historical mean values.
Luis E. Cervera Gómez
Nóesis
138
Impacts of Droughts on the Rio Conchos Watershed over the International
Water Treaty Between Mexico and the United States of America
References:
Climatic Prediction Center. “Weekly Palmer Drought and Crop Moisture Data
Products Explanation”. 2000 [Online] Available in http://www.cpc.ncep.
noaa.gov/products/analysis_monitoring/cdus/palmer_drougt.
Comisión Nacional del Agua (CNA). “Programa Hidráulico de Gran Visión del
Estado de Chihuahua 1996-2020”. Subdirección General de Programación,
Gerencia Regional Norte, Gerencia Estatal de Chihuahua. 1997, Contrato:
SGC-GRN-CHIH-96-82.
Fierro, L.C. “Relación de sequías en la antigua Nueva Vizcaya y en el actual
estado de Chihuahua”. 1999 [Online] Available in http://www.sequia.edu.
mx/articulos/lcf-01.html.
IBWC. “Deliveries of Waters Allotted to the United States Under Article 4 of
the United States-Mexico Water Treaty of 1944”, Report of the United Sates
Section, 2002.
IBWC. “Utilization of Waters Treaty of 1994. Treaty Between the United
States of America and Mexico”, 2002 Available in http://wwwibwc.state.gov
[2/23/2002].
INEGI. “XII Censo General de Poblacion y Vivienda: Resultados Preliminares”,
2000, [Online] Available: http://www.inegi.gob.mx/difusion/espanol/cpyv/
indice.html.
Kelly, M.E. “The Rio Conchos: A preliminary Overview”, Texas Center for
Policy Studies, 2001.
Steila, D. “Drought in Arizona: A Drought Identification Methodology and
Analysis”, University of Arizona, Tucson, Arizona, 1972.
The National Commission for the Knowledge and Use of Biodiversity
(CONABIO), “Cuenca Alta del Río Conchos y Río Florido”. 2001, [Online]
Available: http://www.conabio.gob.mx.
USDOI. “Water-Resources in the Rio Grande-Rio Conchos to Amistad
Reservoir Subarea”. Field Coordinating Committee, Fact Sheet 3, April
1988.
