ABSTRACT Climate change is already affecting life on Earth and mitigation measures must be taken, which require proper identification of the regions most vulnerable to these changes. This study assesses the impact of climate change on water availability (water surplus, water deficit and precipitation minus evapotranspiration) and climate classification across the Caatinga biome, located within the Brazilian semiarid region. To obtain the water surplus and water deficit, the water balance of Thornthwaite & Matther (1955) was applied considering the current scenario (historical series from 1961 to 2015); the optimistic scenario, considering the Representative Concentration Pathway (RCP) for a radiative forcing of 4.5 W.m-2 (RCP4.5), i.e., considering a temperature increase of 1.5 ºC and a reduction in rainfall from 15%; and pessimistic, considering a radiative forcing of 8.5 W.m-2 (RCP8.5), i.e. taking an increase in temperature of 4.0 ºC and a reduction in rainfall from 20%. The results showed that the Caatinga was strongly impacted, especially in the pessimistic scenario, with no water surplus and a negative water balance. For the optimistic scenario, only a small portion presented a water-surplus and a positive water balance. Projections also showed that a large part of the Caatinga changed from semiarid to arid climate.

Melon plants (Cucumis melo L.) are grown in the state of Rio Grande do Norte (RN), the largest producer of melons in Brazil, with plastic mulch and agrotextiles. Studies of crop evapotranspiration (ET) under these conditions are required to ensure adequate irrigation. This study aimed to determine the crop coefficients (Kc) of irrigated melon plants grown with mulch and agrotextiles in the region of Mossoró, RN, based on the Bowen Ratio Energy Balance (BREB) method. Two experiments were conducted at different times during the 2009/2010 season in a melon producing area (4°59'52" S, 37°23'09" W, and 54 m elevation) to define ET and Kc. Due to the plastic cover and reduced precipitation during the experiments, the Kc obtained by the BREB method was considered the basal Kc - KcbBREB. The results were compared with the Kcb from the FAO 56 Bulletin. There was close agreement between BREB and FAO ET measurements (12 % underestimation by the FAO method for the entire crop season), with sizeable differences only during the initial phenological stage. The mean KcbBREB values of the two field campaigns were 0.26, 0.96 and 0.63 for the initial, midseason and late stages, respectively. The high KcbBREB value in the initial growth phase may be related to the effect of the plastic mulch and agrotextiles on the energy balance at the surface. The relationship between KcbBREB and fc had high correlation, making possible an estimation of the melon Kcb based on the level of crop ground cover.

O objetivo desse estudo foi avaliar um método simplificado, baseado em classes de NDVI para refinamento das imagens de temperatura da superfície (Ts), obtidas pelo sensor TM do Landsat 5 referentes aos anos de 2005 e 2006. Para tanto, foram propostos e comparados três modelos de refinamento baseados no método de regressão linear. Os erros percentuais e erros médios quadráticos obtidos com a utilização dos modelos avaliados foram, respectivamente, da ordem de 0,37% e 1,38 ºC, enquanto o modelo original apresentou erro médio quadrático da ordem de 1,32 ºC. Foram constatados que os erros obtidos com as calibrações realizadas não influenciaram significativamente nos valores médios das imagens termais, e que os resultados contribuíram substancialmente para a melhoria da resolução espacial das mesmas. O refinamento permitiu ainda a identificação precisa de alvos da superfície e a identificação de feições não detectáveis na resolução original. Isto evidencia que o método simplificado sugerido neste estudo, permite um refinamento preciso com uma forma de obtenção mais simples em relação ao modelo original.

The objective of this study was to use a simplified method based on NDVI classes for the sharpening of the Landsat 5 - TM surface temperature images (Ts) obtained during the years of 2005 and 2006. Thus, three sharpening models, based on the linear regression method, were proposed and compared. The relative and the root mean square errors obtained through the suggested models were of 0.37% and 1.38 ºC, respectively, while the original model presented root mean square error of 1.32 ºC. It was verified that the errors obtained with the accomplished calibrations did not significantly influence in the average values of the thermal images and the results contributed substantially to the improvement of their spatial resolution. The sharpening allowed the precise identification of the targets and features undetectable at the original spatial resolution. This evidences that the simplified method, suggested in this study, allows an accurate sharpening more easily applicable than the original model.

O presente estudo teve como objetivos estimar a evapotranspiração - ETc e determinar a curva do coeficiente de cultura - Kc do algodoeiro irrigado através do Surface Energy Balance Algorithm for Land - SEBAL com imagens orbitais TM - Landsat 5. Foram utilizadas oito imagens distribuídas ao longo do ciclo fenológico do algodoeiro cultivado na Fazenda Busato localiza no município de Bom Jesus da Lapa, região do Médio São Francisco, Estado da Bahia (13°15'18''S, 43°25'05''W, 436 m). A classificação climática da região segundo Köppen é BSwh'. O saldo de radiação foi calculado a partir de imagens da temperatura, emissividade da superfície, índices de vegetação, albedo e calculados os fluxos de calor no solo e sensível para obter-se o fluxo de calor latente e a ETc. Verificou-se que o índice de vegetação NDVI apresentou evolução concomitante com o ciclo da cultura, com valores máximos (0,80) aos 70 dias após semeadura - DAS. A ETc e o Kc obtidos foram, respectivamente: 1,0 a 5,0 mm dia-1 e 0,65 no período de desenvolvimento (7 e 70 DAS); > 6 mm dia-1 e 1,18 durante a floração e formação dos capulhos e 2 mm dia-1 e 0,66 no fim do ciclo. Os resultados mostram que o NDVI é um bom indicador do desenvolvimento do algodoeiro e os dados de ETc e Kc estão coerentes com relatos na literatura.

This research aimed determine cotton evapotranspiration - ETc and crop coefficient - Kc slope using the Surface Energy balance Algorithm for Land - SEBAL with TM-Landsat 5 images. We used eight images distributed throughout the cotton growth season on the Busato Farm located in Bom Jesus da Lapa, Médio São Francisco region, Bahia state (13°15'18" S, 43°25'05" W, 436 m). The Climate classification of region by Köppen is BSwh'. The net radiation was calculated from surface temperature, surface emissivity, vegetation index and albedo imagesn and calculated soil and sensible heats fluxes to obtain the latent heat flux and ETc. The NDVI vegetation index presented concurrent development with the crop season, with maximum values (0.80) 70 days after sowing - DAS. Values of Etc and Kc were, respectively, 0 a 5.0 mm day-1 e 0.65 in vegetation development period (7 e 70 DAS); > 6 mm day-1 e 1.18 during flowering and cotton bolls generation; and 2 mm day-1 e 0.66 in the crop cycle ending. Results show that NDVI is a good indicator of the development of cotton and ETc and Kc data are consistent with reports in the literature.

O sensoriamento remoto tem se mostrado eficaz na avaliação de fluxos de energia e de propriedades biofísicas de superfícies vegetadas em escala regional. No presente trabalho, utilizou-se o algoritmo SEBAL - Surface Energy Balance Algorithm for Land e imagens TM - Landsat 5 para mapeamento e quantificação do albedo (α), NDVI, temperatura da superfície (Ts) e radiação líquida (Rn) em área de algodão irrigado por pivô central, na Fazenda Busato (13,25º S; 43,42º W; 436 m), município de Bom Jesus da Lapa, Bahia. Seis imagens de céu limpo ao longo do período da cultura (janeiro a agosto de 2007) e os respectivos dados meteorológicos foram utilizados para implementação do algoritmo. Após o processamento digital das imagens, verificou-se nítida relação dos parâmetros α, Ts e NDVI com o desenvolvimento da cultura. Os menores valores de α (10 a 20%) e Ts (< 24ºC) e os maiores de NDVI (>0,75) ocorreram na fase de máxima cobertura do solo. A radiação líquida (Rn) diminuiu progressivamente com o tempo, influenciada, principalmente, pela diminuição da radiação solar incidente com o aumento do ângulo zenital. Os valores de Rn variaram de 430 W m-2 a 700 W m-2 nos pivos cultivados. A técnica de sensoriamento empregada capturou de forma nítida a variabilidade temporal e espacial de Rn e dos parâmetros biofísicos, cujos valores encontrados são compatíveis com os reportados na literatura para a mesma cultura sob regime de irrigação.

Remote sensing is currently an important tool for evaluation of net radiation and biophysical parameters over vegetated surfaces on a regional scale. In this research, the SEBAL - Surface Energy Balance Algorithm for Land and TM - Landsat 5 images were used to map and quantify the albedo (α), NDVI, surface temperature (Ts) and net radiation (Rn) of center-pivot irrigated cotton fields in the Busato Farm (13.25º S; 43.42º W; 436 m asl), western of State of Bahia, Brazil. Images from six clear-sky days during the cropping season (January to August 2007) and the corresponding meteorological data were used to run SEBAL. Results showed a clear relationship between α, NDVI, and Ts and crop development. The lowest values of α (10 to 20%) and Ts (< 24ºC), as well as the highest values of NDVI (> 0.75), occurred at the time of maximum ground cover. Net radiation decreased progressively with time, following the decrease of the incident solar radiation with increasing solar zenith angle. In the cropped center pivots, Rn ranged from 430 Wm-2 to 700 Wm-2. The remote sensing technique used in this study was consistent in capturing the temporal and spatial variability of the biophysical parameters and net radiation over the irrigated cotton, and their values are in agreement with those reported in the literature for the same crop in irrigated areas.