«DIPLOMA THESIS Linking Climate Change with Food Security in the highlands of Khyber Pakhtunkhwa, Northwest Pakistan Presented by: Martin Kienzler ...»
If looking at the seasonal progress of precipitation the most distinct feature is that the reduction of rainfall along the monsoon belt happens during the summer months and, at a lesser rate in spring. This matches with the most recent trend of summer precipitation, recorded in the CRU data (see ﬁgure 6.14) and would mean that the intensity of the Indian summer monsoon would weaken further during the next one hundred years. The distinct increase in precipitation along the Karakoram ranges as mentioned above mainly occurs in winter and spring. For the lower lying regions of Pakistan a relatively strong increase is projected for the autumn months, while the monsoon induced summer precipitation will be reduced considerably.
Figure 6.17 – Diﬀerence of the 2071-2100 mean and the 1971-2000 mean for temperature (left, ◦ C per century) and precipitation (right, mm per century) for diﬀerent seasons (from top to bottom: spring, summer, autumn, winter).
REMO model output
6.4 Qualitative research The following subsections contain information, all of which are based on the statements of personal interviews and observations.
6.4.1 Kaghan valley The valley of Kaghan is populated only during summer. With the beginning of winter and the ﬁrst snow, which usually occurs between mid and end of October, the population leaves the valley to spent the winter in the cities south of the valley like Balakot, Mansehra or even Abottabad. The reason for this is the big amounts of snow inside the valley and hence the impossible accessibility during winter. This implicates that in Kaghan valley only one cropping season is possible. The cropping of wheat and therefore the cropping during two seasons is only possible below an altitude of around 4,000 feet (approximately 1,200 masl, SARC (2011)).
The main crops of the upper Kaghan valley are potatoes, green peas and maize. The humble diversity of crops is the consequence of the harsh climate in elevations of distinctly above 2,000 masl. Some farmers in the lower parts of the valley cultivate additional products like beans, walnuts, apples, peaches and vegetables. Because of climate, the potatoes seem to be secure from every type of disease and therefore are exported to all over the country. Sowing usually begins during the ﬁrst week of May and is ﬁnished by the 15th of June. Harvesting occurs between the end of September and the beginning of October. All of the informants reported a later onset of winter (which is deﬁned as the time when the population leaves the valley) and a shorter duration of the time they have to spent out of the valley. This implicates a strong warming of the winter months and theoretically means a longer vegetation period available for the crops.
Apart from this observation, the amount of snow during winter is reported to be considerable less, and most people estimated it now to be only the half of the amount during the 1970s. This could be either due to higher temperatures or because of less winter precipitation. Furthermore all the interviewed persons reported distinctly higher temperatures during summers, during daytime as well as at night. The statements concerning maximum (daytime) temperatures however were more distinct. Most people referred the warming of winter and summer temperatures to the strong deforestation which still is occurring inside the valleys and in the surrounding areas. Rainfall amounts during the summer months are rather said to be less than around twenty to thirty years before.
Especially the year 2011 was extremely hot and dry. Some people reported that they observe drought periods more frequently.
The yields of maize are estimated to be around 1,300 kg per acre. The yields of wheat in the lower regions are supposed to be between 500 and 1,000 kg per acre, whereas they add up to around 3,500 to 4,000 kg per acre in the plain areas of Pakistan (SARC, 2011).
This points out that yields are considerably less in mountain regions than in plain areas.
Almost all of the questioned persons reported increasing yields over the last twenty years.
But however they all made chemical fertilizer responsible for that, which they use since the late eighties.
Chapter 6. Results
An interesting observation is that during the 1970s only around 30% of the arable land was cultivated. Until now this percentage reached almost 70%, which is said to be due to the better road conditions and the bigger demand of the markets in the cities south of the valley. Furthermore in the meantime most people have found additional sources of income in the tourist sector for example as hotel or restaurant keeper, jeep driver, guide, etc. This in contrast implicates that agriculture somehow looses its strong importance.
Table 6.1 – Key results of the semi-structured interviews.
++ means strong increase, + slight increase, – slight reduction, – – strong reduction and 0 no trend 6.4.2 Chitral district The variety of agricultural products is a lot higher in the valleys of Chitral district than in Kaghan valley. This is due to lesser elevation, higher winter temperatures, less precipitation amounts and hence less snow during winter. Therefore two cropping seasons are possible. This is the most remarkable diﬀerence to agriculture in Kaghan valley. Main crops in this part of Khyber Pakhtunkhwa are wheat, maize, rice, barley, millet (bajra) and sorghum. Furthermore beans, peas, lentils, potatoes and a variety of vegetables are
Chapter 6. Results
cultivated. As well a lot of diﬀerent fruit trees, namely walnut, apple, apricot, pear, cherry and peach supplement the agricultural diversity.
This high variability of crops however is enabled through irrigation, which is necessary due to the low summer precipitation. Before the development of irrigation channels in most regions crops could be grown only during the rabi (winter) season. In some villages (e.g. Mastuj) irrigation channels were not built until 1985.
The kharif season approximately corresponds to the cropping season in Kaghan valley.
But sowing and harvesting dates depend on the diﬀerent crops. Lentils and beans are grown between April and late July. Maize is sown in May and harvested in October, fodder Maize is sown in July because it does not have to reach maturity. Rice generally is sown between the ﬁrst and the tenth of July and harvested during the last week of October. In the lower parts of the district around Ayun the whole season is shifted to earlier dates around two to three weeks. The rabi season is the season of wheat and barley. Sowing occurs in late October or early November, while crops are harvested by the end of June or beginning of July. Most persons did not recognize a trend of changing sowing or harvesting dates but reported a general ﬂuctuation of about ten to ﬁfteen days, depending on the weather of the respective year. Only in Mastuj some informants said that the sowing generally happens earlier now, but they related this fact to the use of machines.
The questioning concerning the developments of yields led to diﬀerent results and showed the strong dependence of fertilizer and irrigation. All the farmers are using chemical fertilizer since the mid or the end of the eighties. Some people reported distinctly increasing yields since then, in some places yields are said to be twice as much as before using fertilizer. Only one informant related increasing yields to climatic changes. Other farmers recognized that their yields began to decrease at some point due to fertilizing.
According to them, the usage of chemical fertilizer strongly aﬀects the fertility of soils, the quailty of crops and the occurrence of diseases.
On the basis of these observations no statements can be made about climatic changes.
The questions directly dealing with climate issues unfold the following. A distinct (slight) temperature increase during the winter months was recognized in Mastuj and Ayun (Garam Chashma). The people in the Kalash valley reported rather colder winters in the last twenty years and a warming trend only during the last few years. Concerning snow amounts all the persons agreed with each other: Amounts of snow are reduced clearly to less of the half of the amount around thirty years ago. Instead of three to four feet now only one to one and a half feet of snow occur. Also the duration of snow cover is reduced to the period December to February (instead of November to March before). Between 2000 and 2005 there seemed to be no snow at all in some valleys. The interviewed persons in the Kalash valleys referred this to rather drier winter months.
The statements concerning changes in summers were more clear. Almost all the informants reported only slightly warmer summer months than thirty years ago. Interesting is, that the night (minimum) temperatures during the summer months generally are said to have become lower in recent years. Some farmers even reported more frost nights.
This comes along with a reduction of the anyway humble summer precipitation amounts.
Furthermore a considerable amount of informants reported increasing events of several
Chapter 6. Results
days without rain as well as heavy rain events. This implicates a trend to more extreme events.
Irrigation now enables to conduct agriculture for more people. The limited space within the narrow valleys together with increasing population pressure forces the farmers to move to higher altitudes, where climate becomes the limiting factor again. Another consequence is that people do not have enough space to cultivate considerable amounts for selling their products. Often their yields even do not suﬃce to meet their own demands.
This is the reason why most of the farmers practice additional jobs. A big problem is moreover the still swindling tourism because of the political situation in the border region to Afghanistan.
7 Discussion Comparison of the datasets CRU and REMO “The South Asian region is considered to be one of the most challenging regions for the climate-modelling community due to its geography, orientation, population, and complexities of processes involved in the evolution of its climate.” (Saeed et al., 2011).
Nevertheless several regional climate models try to simulate the climatology of South Asia, one of them being REMO, which is used for analysis in this study. This section shall compare the output of REMO with the observational dataset CRU and evaluate which of the datasets is more useful to make statements concerning climate change in this region. Subsequently the output is compared with the data of the meteorological stations for validation.
The analysis of the two gridded datasets has shown that they do not agree in every aspect. When calculating mean climatology, regarding temperature as well as precipitation they suit with each other quite good. It becomes obvious that REMO is capable to simulate the climatological patterns of the region very well. The only diﬀerence is that the REMO simulations overestimate the temperatures in relatively warm areas, like the Indus plains and in contrast underestimate the temperatures in the high mountain areas, compared to the CRU observations (compare Mannig et al. (2012)). The warm bias of the REMO simulation over the plain areas in the range of the thermal heat low was also recognized by Saeed et al. (2009). The same results apply for precipitation. For arid regions like the southwestern part of Pakistan REMO simulates a distinctly lesser precipitation as recorded by CRU. For the high precipitation belt within the area of inﬂuence of Indian summer monsoon REMO suggests higher precipitation values than CRU. The biases shown by the REMO simulations is not unusual: According to Pal et al. (2007) and Ashfaq et al. (2009) the for Asia widely used regional climate model RegCM3 showed the same characteristics, for example a warm bias over the semidesert and desert regions of southeastern Pakistan and a positive precipitation bias over monsoon regions.
But anyway REMO takes into account high mountain areas, where no meteorological stations are available, like the mountains of Belochistan, the Sulaiman range or the Karakoram range along the Chinese border. For these regions the CRU data interpolates the data from the nearest available station, which often is a few hundred kilometres apart and represents a totally diﬀerent climate. This often leads to a considerable bias for the respective grid boxes (Mitchell & Jones, 2005; Paxian et al., 2011). Especially for these regions the REMO data simulates a more realistic climatology. Along with this the decisive advantage of the REMO output is the higher resolution, which allows statements on a far smaller scale. For example the narrow valleys of Himalaya and Hindu Kush can
Chapter 7. Discussion
be distinguished from the surrounding high mountain areas, which is quite important for this study. These areas are however not represented by the CRU dataset.
On the other hand there is a big diﬀerence between the results concerning temperature and precipitation when regarding climate trends. For temperature both datasets show similar tendencies on a large scale: a positive trend for most regions during autumn and winter and a rather cooling tendency during spring and summer. But the most noticeable diﬀerence is the strong dipole visible in the CRU data between the region around Parachinar and the area west of it, which is described more exactly in section 6.1.2. This seems not to be very realistic and could be based on a bias in the recorded data of one of the meteorological stations, of which the CRU data is interpolated from.
An analysis of the data of the climate station Parachinar in fact has shown a strong negative trend, too, but the question is how reliable the records are. The negative trend in this area is conﬁrmed by the REMO simulations only in spring.