Observed changes in climatic variables relevant for agricultural production during in the Pampas region of Argentina, Brazil and Uruguay (AIACC Project No. LA27) Walter E. Baethgen (IRI Columbia University-INIA, Uruguay) Agustin Gimenez (INIA, Uruguay) Graciela Magrin (INTA-Argentina) Maria I. Travasso (INTA-Argentina) Gilberto Cunha (EMBRAPA, Brasil) Mauricio Fernandes (EMBRAPA, Brasil)
Objectives Explore changes in the last years observed in climate variables that are relevant for agricultural production Establish the basis for one of the methods used in our project to create climate change scenarios: project changes observed in the last years to the next years using a weather generator (G. Magrin’s presentation)
Methods Linear regression of mean trimester rainfall and temperatures for the entire available period (1930 – 2000) Stat significance using parametric and non-parametric (Kendall) tests Plot slopes using a contouring and surface mapping program (Surfer) to study spatial distribution of changes Expected Result Annual mean changes of the trimester means for Rainfall, T Max, T Min in the entire period ( ) Remaining Question: Were changes uniform throughout the entire period? Time
Methods (2) Compare rainfall and temperature means for different periods: 10-year periods: , , …, year periods: , , …, year periods: vs
RESULTS Changes in Rainfall (mm Trimester / year) Linear regression Analysis
Changes in Rainfall (mm Trimester / year)
Changes in Rainfall (mm Trimester / year) Linear regression Analysis Significant Slopes (P<0.10)
Changes in Rainfall (mm Trimester / year)
Summary by Trimester
Fall
Winter
Spring
Summer
Comparing two Periods: vs
Changes in Temperatures Mean monthly T Max Mean monthly T Min Regression Analysis
JFM T Min MJJ JFM JJA T Max
Comparing two Periods: vs
T Max: vs (Summer )
T Max: vs (Summer )
T Min: vs (Winter )
T Min: vs (Winter )
T Min: vs (Summer )
T Min: vs (Summer )
Relevance for Agricultural Production: Frost Regime
UY = Estanzuela AR = Pergamino
Changes in frost regime estimated with regressions ( )
2000 compared to 1930: Frost period starts later and ends earlier Less number of frosts in the frost period Mean temperature of frosts is higher
Changes in absolute and in mean TMax and Tmin Estimated with Regressions (1930 – 2000)
Changes in absolute and in mean TMax and Tmin Estimated with Regressions (1930 – 2000) 2000 compared to 1930: Absolute TMax and Mean TMax are lower Absolute TMin and Mean TMin are higher
Reported Changes in Means Changes in Variability?
7+, 2-, 1= 3+, 3-, 4= 4+, 4-, 2= 5+, 5=
Conclusions Increased Rainfall in Spring and Summer Better for pastures, summer crops but animal / crop diseases? pests? (M. Fernandes’ presentation) Increased Rainfall Variability in Summer (?) (higher risk) Decreased Mean T Max (Spring and Summer) Increased Mean T Min (all year) Shorter frost period (first – last dates) Less number of days with frost Frosts became less severe (Temp is higher)