Drought Monitoring

Currently drought monitoring in Mexico is in its early stage of development. There are only two institutions that monitor droughts in the country. One of them is the SMN, and the other is the Centro de Investigaciones Sobre la Sequía (CEISS; i.e., Drought Research Center). The SMN monitors droughts by determining precipitation deviations from the normal, and these deviations are regularly published (http://smn.cna.gob.mx/productos/ map-lluv/tabla.gif). Essentially, the historical monthly mean and annual mean are compared with the actual monthly and annual precipitation; the deviation is presented as a percentage of mean values. A negative deviation represents precipitation deficit (dry condition or drought), while a positive one indicates surplus (humid condition or wet period).

Recent cooperation efforts among the Mexican Comisión Nacional del Agua, the U.S. National Oceanic and Atmospheric Administration, and the Meteorological Service of Canada have produced the North American Drought Monitor, which presents a monthly drought map of southern Canada, the United States, and Mexico. The drought monitor is the result of a process that synthesizes multiple indices, outlooks, and local impacts into an assessment that best represents current drought conditions. The final outcome of each drought monitor is a consensus of federal, state, and academic scientists (NCDC, 2003). An advantage of this small scale is that the climatic phenomena can be visualized in all its extension; a disadvantage is that the scale is not practical for important drought-affected areas.

The Mexican Drought Research Center (CEISS) applies the standardized precipitation index (SPI; chapter 9; McKee, 1993, 1995). Total monthly precipitation data recorded by the SMN in Chihuahua State is obtained by the CEISS during the first week of each month. The data are processed using specially designed software to compute the monthly SPI. The software stores the state historical precipitation database since 1970, which is automatically updated every month. SPI values are computed on an interval of three months (SPI-3) and 12 months (SPI-12). The SPI-3 is intended to help dry-land farming where production depends on seasonal rainfall, whereas SPI-12 can be used by federal and state agencies that manage state water resources. Finally, the SPI values are loaded on to a Geographic Information System (GIS) for showing spatial distribution of droughts and their severity levels (http://www.sequia.edu.mx/). Further GIS analyses are performed to estimate damages to agricultural areas, other natural resources, and society.

The advantage of using a large scale (state size) is that regional resources affected by drought can be better evaluated. A main disadvantage, however, is that the geographic limits of major drought events affecting large areas might not be included in the drought monitor. The current monitoring activities of the CEISS are confined to the state of Chihuahua. The center has plans to extend its monitoring activities to the northern part of Mexico. The SMN is also planning drought monitoring at the national level by applying the SPI methodology in addition to the precipitation deviation method described earlier.

Predicting Crop Yields using Standardized Precipitation Index

The SPI is a qualitative indicator of drought. To use it for quantitative assessment of agricultural drought, the quantitative relationship between SPI and yields of maize and bean crops was examined for the north-central states of Mexico. Table 10.2 shows the basic variation in main yields of maize and bean during the last two decades.

The rainfall data were collected from two databases (ERIC I, 1995; ERIC II, 1999) of the Instituto Mexicano de Tecnología del Agua (IMTA; i.e., Mexican Institute of Water Technology) to compute SPI-3. In addition, maize and bean yield data were collected from the Sistema Integral de Información Agroalimentaria y Pesquera (SIIAP; i.e., Agricultural and Food Information Integral System) created by the federal agency, Secretaría de Agricultura, Ganadería, Desarrollo Rural, Pesca y Alimentación (i.e., Agency of Agriculture, Cattle, Rural Development, Fishing and Food; SIIAP, 2003) and INEGI (1993,1995,1998,1999a, 1999b, 2000b, 2000c).

Results of these correlations are presented in figure 10.4. The results show positive but low correlation coefficients (from 0.26 to 0.54) for bean and maize yields with SPI-3. In addition to SPI-3, the correlation analysis was attempted using SPI-4, SPI-5, and SPI-6, but the results showed even lower correlations. Aguascalientes, Durango, and Zacatecas are the states that indicate moderate correlations for both crops. Correlations for Chihuahua and Coahuila are the lowest (figure 10.4b,c). SPI-3 seasonal values for the state of Durango show the best correlation with the economic values of bean and maize crops.

A possible reason for low correlations is that the SPI values used in the correlation analysis were averaged for an entire state. One way to improve these results is by splitting the state into regional agricultural areas and

Table 10.2 Basic statistics showing the variation in bean and maize yields in the five north-central states of Mexico

State

Bean (tons/ha)

Maize (tons/ha)

Min

Max

Mean

Std. dev.

Min

Max

Mean

Std. dev.

Aguascalientes

0.06

0.36

0.17

0.07

0.17

0.93

0.36

0.18

Coahuila

0.20

0.60

0.40

0.09

0.33

0.91

0.62

0.14

Chihuahua

0.20

0.73

0.40

0.12

0.13

1.15

0.75

0.24

Durango

0.17

0.67

0.40

0.16

0.38

0.95

0.64

0.14

Zacatecas

0.14

0.57

0.39

0.12

0.50

1.02

0.68

0.13

National

0.32

0.57

0.47

0.07

1.39

2.05

1.68

0.16

carry out correlation analysis between regional yields and corresponding SPI values. Another reason for poor relationship is that SPI depends only on rainfall, which is just one of the several variables that affect a crop yield. Though a simple indicator, SPI cannot be used to satisfactorily predict crop yield and therefore cannot be used for quantitative assessment of agricultural droughts in Mexico. There is a need to develop an agricultural drought index that can be used to predict agricultural drought by predicting yields of major crops.

0 0

Post a comment