Palfai Drought Index (PaDI) Easy method to analyze drought, tool for forecast, for early warning DMCSEE Final meeting 14 th May 16 th May 2012, Ljubljana Arpad Herceg ATI-VIZIG, Szeged, Hungary hercega@ativizig.hu
PaDI (Palfai Drought Index) Indicates strength of drought for one agricultural year with one data Proportional with crop decrease Uses results of meteorological measures, but the summer precipitation is more weighted characterization of water scarcity in agriculture Contains the most determinant factors Gives time series proper for trend analysis Suitable for correlation analysis (yield, pests, etc.)
PaDI (Palfai Drought Index) Based on Hungarian PAI - Theoretically the same, but less data demand - Calculating is more simple - Relates to local average conditions Calculation data demand: - monthly mean temperature - monthly precipitation sum Possibility to preparing entire map of droughtness (PaDI 10% ) for larger regions (countries, Carpathian Basin, SEE, Europe)
PaDI (Palfai Drought Index) PaDI o PaDI = aug T i= apr = sept 10 PDI + i ( P ) i * wi i= oct 5*100 o * k1 * k 2* k 3 Month w i X 0.1 XI, XII 0.4 I,II,III,IV 0.5 V 0.8 VI 1.2 VII 1.6 VIII 0.9 IX 0.1 Σ= 7.5 k 1 k 2 = relation between summer average temperature and annual average = relation between summer min. monthly precipiation and annual average k 3 = relation between average precipiation of previous 36 month and annual average
PaDI (Palfai Drought Index) 1994 Evaluation of PaDI ( for a year) 4-6 mild drought 6-8 moderate drought 8-10 medium-weight drought 10-15 serious drought 15-30 very serious drought >30 extreme drought
PaDI (Palfai Drought Index) 1979 1993 1988 1990 1994 2000 Evaluation of PaDI ( for a year) 2003 2007 4-6 mild drought 6-8 moderate drought 8-10 medium-weight drought 10-15 serious drought 15-30 very serious drought >30 extreme drought
PaDI (Palfai Drought Index) Regional averages of PaDI in period 1961-2009
Regression between PaDI and PDSI 16.00 PAI, C/100mm PAI, C/100mm 16.00 14.00 Kecskemét 14.00 Debrecen 12.00 12.00 10.00 10.00 8.00 8.00 y = 4.9949e -0.1648x R 2 = 0.7312 y = 4.1149e -0.1499x R 2 = 0.7189 6.00 6.00 4.00 4.00 2.00 2.00 0.00 PDSIaug 0.00 PDSIaug -6-4 -2 0 2 4 6-6 -4-2 0 2 4 6 in Hungary y = a * e bx ; R 2 avg >0,7
Regression between PaDI and scpdsi 20 PaDI 20 PaDI 18 y = 3,102e -0,1499x R 2 = 0,8193 18 y = 3,3395e -0,1265x R 2 = 0,7862 16 VARAZDIN (HR) 16 BJELOVAR (HR) 14 Expon. (VARAZDIN (HR)) 14 Expon. (BJELOVAR (HR)) 12 12 10 10 8 8 6 6 4 4 2 2 0-10 -8-6 -4-2 0 2 4 6 8 10 scpdsi 0-10 -8-6 -4-2 0 2 4 6 8 10 scpdsi Croatia y = a * e bx ; a avg = 4,05 b avg =-0,13 R 2 avg = 0,69
Regression between PaDI and scpdsi 20 PaDI 20 PaDI 18 18 16 y = 15,679e -0,2129x R 2 = 0,8157 16 14 14 y = 16,363e -0,1466x R 2 = 0,7374 12 12 10 10 8 8 6 6 4 ATHEN (GR) 4 HELLINIKON (GR) 2 Expon. (ATHEN (GR)) 2 Expon. (HELLINIKON (GR)) 0-10 -8-6 -4-2 0 2 4 6 8 10 scpdsi 0-10 -8-6 -4-2 0 2 4 6 8 10 scpdsi Greece y = a * e bx ; a avg = 13,10 b avg =-0,13 R 2 avg = 0,69
Regression between PaDI and scpdsi 20 PaDI 20 PaDI 18 16 y = 3,4503e -0,1206x R 2 = 0,7666 18 16 y = 3,6544e -0,1733x R 2 = 0,8529 14 14 12 NAGYKANIZSA (HU) Expon. (NAGYKANIZSA (HU)) 12 SZOMBATHELY (HU) Expon. (SZOMBATHELY (HU)) 10 10 8 8 6 6 4 4 2 2 0-10 -8-6 -4-2 0 2 4 6 8 10 scpdsi 0-10 -8-6 -4-2 0 2 4 6 8 10 scpdsi Hungary y = a * e bx ; a avg = 4,40 b avg =-0,14 R 2 avg = 0,71
Regression between PaDI and scpdsi 20 PaDI 20 PaDI 18 18 16 y = 3,0637e -0,1459x R 2 = 0,7982 16 y = 2,9552e -0,1741x R 2 = 0,874 14 14 LENDAVA (SI) VELIKI DOLENCI (SI) 12 Expon. (LENDAVA (SI)) 12 Expon. (VELIKI DOLENCI (SI)) 10 10 8 8 6 6 4 4 2 2 0-10 -8-6 -4-2 0 2 4 6 8 10 scpdsi 0-10 -8-6 -4-2 0 2 4 6 8 10 scpdsi Slovenia y = a * e bx ; a avg = 2,55 b avg =-0,14 R 2 avg = 0,72
Regression between PaDI and maize yield Maize annual average yield, t/ha (1990-2009) in Hungary, R 2 =0,78 PaDI annual average, ºC/100mm (1990-2009)
Regression between PaDI and maize yield Maize annual average yield, t/ha (1990-2009) Bács-Kiskun county, R 2 =0,95 PaDI annual average, ºC/100mm (1990-2009)
PaDI 10% (Palfai Drought Index - frequency) Drought frequency, map of droughtness: - 10% probability of occurrence (log-pearson Type III Distribution) of PaDI, given from long data queue - expresses the climatic factor (geographical characteristic) of region drought intensity and frequency,
PaDI 10% (Palfai Drought Index - frequency) The 10% probability of occurrence of PaDI for SEE region (PaDI 10% ) PaDI 10% Drougtness Drougtness Probability of impacts (Risk * Impacts = Vulnerability)
PaDI forecast (early warning) PaDI 2007
PaDI forecast (early warning) Drought forecast in May 2007 In case of dry weather In case of average weather
PaDI forecast (early warning) Drought forecast in June 2007 In case of dry weather In case of average weather
PaDI forecast (early warning) Drought index in 2007
PaDI forecast (early warning) Drought forecast in April 2012 In case of dry weather In case of average weather
PaDI forecast (early warning) Drought forecast in May 2012 In case of dry weather In case of average weather
PaDI (Palfai Drought Index) Conclusions, suggestions: - Calculation, data processing and understanding of PaDI is easy - Expresses well crop decrease - Correlation between scpdsi and PaDI is good (R 2 >0,7) - PaDI and SPI 3 (for spring or summer) together expresses strength and reason of drought - Useful for medium-term forecast of drought - Practical for multi-decade data series analysis and mapping, for the assessment of the possible climate change scenarios regarding drought.
PaDI (Palfai Drought Index) Thank you for your attention! Arpad Herceg ATI-VIZIG, Szeged, Hungary