1. Introduction
Afforestation has become an important option for reducing the negative effects of anthropogenic climate change. As a result of the Clean Development Mechanism of the Kyoto Protocol and the Certified Emission Reduction, it became an alternative concept for carbon capture and storage. Afforestation can become economically interesting for companies, which are large emitters of carbon dioxide. In the following, we define the economic application of afforestation as carbon farming and the analysis of its impact on weather and climate as climate engineering by afforestation (CEA). Afforestation projects are biogeoengineering efforts that, when implemented through mono/mixed forestry/agroforestry systems or managed timber forests, could potentially offer multiple benefits such as 1) carbon sequestration in vegetation and soil carbon stocks, 2) carbon trading under national and international agreements, 3) commercial crop value by production of oils and energy production, 4) environmental services such as soil protection and water filtering, and 5) mesoscale climate modification by changes in land surface properties. Because of its carbon binding potential, CEA is considered a mitigation option within the Intergovernmental Panel on Climate Change (IPCC; Metz et al. 2007).
It is interesting to compare the impact of afforestation with other geoengineering approaches currently under consideration. An overview is given, for example, in Lenton and Vaughan (2009). Geoengineering options are characterized by their impact on the top-of-the-atmosphere radiation balance and on the surface energy budget, which both translate into a radiative forcing potential. While conceding that this may provide a first insight, this approach disregards important feedback processes in the Earth system (e.g., with respect to regional changes in evapotranspiration and the mesoscale water cycle).
Thus, anthropogenic modifications to Earth’s energy balance (e.g., from land-use change or CO2 emissions) would not simply be reversed by geoengineering methods. To take just one example of aerosol injection, Haywood et al. (2013) suggest that three of the four driest Sahelian summers were preceded by substantial Northern Hemisphere volcanic eruptions. In other words, apart from any potential global forcing induced by CO2 sequestration or changes in atmospheric albedo, the possibility of localized and synergistic effects must also be considered. This renders the outcome from all geoengineering methods highly uncertain. Extreme care therefore needs to be taken in the analysis of the net effects on the Earth system, both globally and regionally. The IPCC states that areas of concern are (i) effects on the global or regional climate system including the land, atmosphere, and ocean and (ii) that significant effects would span national boundaries (Edenhofer et al. 2012). This requires conducting a comprehensive assessment, if this is indeed possible, of induced feedback mechanisms in water, matter, and biogeochemical cycles, starting at Earth’s surface.
Large-scale afforestation strategies need to be assessed from a holistic perspective, particularly in regard to multiscale impacts and threats to food security. Glenn et al. (1992) highlighted a fundamental problem with reforestation: “much of the same land required to grow trees may be needed for food production and other uses as the world’s population continues to grow.” They discussed a potential for halophytes, or salt resistant species, to be planted in arid regions as a climate change mitigation strategy. Afforestation could therefore take place in hot and dry coastal areas, in order to avoid competition with other land uses. Nowadays, robust perennial plants are better understood with respect to their physiological properties and carbon binding potential (Fairless 2007; Righelato and Spracklen 2007). For instance, Becker et al. (2013) demonstrated that Jatropha curcas could capture 17–25 tons of carbon dioxide per hectare per year from the atmosphere (averaged over 20 yr). The same progress holds for key technologies such as the setup and operation of desalination plants, which eliminates the use of fossil water for irrigation. A holistic, transdisciplinary approach is necessary (McAlpine et al. 2010). A corresponding study of suitable land areas, carbon farming, and CEA with respect to its economic value is presented in Becker et al. (2013).
Another key topic concerning plantations in desert regions are potential influences on regional weather and climate. Because of the decrease of the surface albedo in combination with a change of the partitioning of the fluxes in the energy balance closure (EBC), a modification to the local circulation can be expected. If such plantations induced net beneficial changes in local climate such as more moderate surface air temperatures and increased cloud formation and rainfall, this would add to the attractiveness of carbon farming projects. Such changes must be thoroughly investigated though at different time and spatial scales before any strong statements can be made, however. Storage of water by a localization of the water cycle through the formation of dew and precipitation would increase the efficiency of water usage in the carbon farming concept. In this work, this modification of the water and energy cycle is studied in two regions in detail, explaining the processes leading to the changes of precipitation induced by plantations in two study areas, the Oman and the Sonora, as shown in Becker et al. (2013).
A careful consideration of geoengineering options requires the application of Earth system models on regional scales with detailed description of the feedback between land surface, vegetation, and atmospheric processes. In our study, we apply an advanced model system of this kind to study CEA with unprecedented detail providing spatially resolved quantitative results. We present climatological mean values and variances of diurnal cycles as well as impacts of jatropha plantations on land surface and atmospheric key variables. In particular, we study the mechanisms leading to convection initiation over the planted area, which in turn lead to increased cloud coverage and precipitation. Understanding these processes is essential for a future, sustainable application of CEA.
2. Climate feedback mechanisms in desert regions
a. Overview
The climate and feedback mechanisms in desert regions have been subject of extensive research (for a general overview see, e.g., Mahmood et al. 2014). Pioneered by Charney (1975), self-stabilization effects in subtropical deserts were related to feedback processes between surface albedo and precipitation. Prentice et al. (1992) considered the feedback processes between vegetation and the hydrological cycle as well. By coupling a climate model to a dynamical vegetation model (Claussen 1994), as well as through conceptual models (Brovkin et al. 1998), two stable regimes were found in the Sahara: a desert equilibrium with low precipitation and absent vegetation as well as a green equilibrium with moderate precipitation and permanent vegetation cover.
Kleidon et al. (2000) studied different climates of a green planet and a desert world with a global climate model and found that the surface temperature was mainly controlled by changes in the hydrological cycle. In spite of the lower albedo of a forested planet, the green planet would have a lower mean surface temperature than a desert world. This is mainly due to a modification of the energy balance by evapotranspiration and a change of the water cycle including cloud development.
In subtropical regions, similar albedo–precipitation effects can be expected. We are focusing on the Oman and the Sonora where large areas of unused, barren land are found at the coast. Here, the large-scale forcing is determined by subsidence caused by the Hadley circulation while horizontal moisture transport is present at least in the summer seasons. In Oman, this is due to humid air advected from the south over the Arabian Sea whereas moist air is transported from the Gulf of Mexico to the Sonora Desert. This is a prerequisite for the induction of precipitation processes, which are studied within this work.
Results based on coarse-resolution climate–biome or conceptual models hint that modified feedback processes over jatropha plantations may bring about a new stable climate regime or change the water cycle (e.g., Charney 1975; Brovkin et al. 1998). Whether these phenomena would actually occur though is uncertain, as these models are relying on a variety of assumptions. These include a coarse representation of the EBC dependent on land surface heterogeneity and vegetation properties. Furthermore, it has been demonstrated that the parameterization of deep convection, which is necessary down to a grid increment of ≈>4 km, led to substantial systematic errors in the simulation of precipitation (Schwitalla et al. 2008; Wulfmeyer et al. 2008, 2011; Rotach et al. 2009a,b; Weusthoff et al. 2010; Bauer et al. 2011). Comparisons between convection-permitting models and models with convection parameterization reveal substantial differences between the soil, cloud, and precipitation feedback processes (Hohenegger et al. 2009; Kendon et al. 2012; Warrach-Sagi et al. 2013).
Therefore, simulations should be performed on the convection-permitting scale without parameterization of deep convection. Convection-permitting resolution optimizes the simulation of land surface exchange, surface heterogeneity, and orographic effects. Recent studies of weather forecast models (Rotach et al. 2009a; Wulfmeyer et al. 2011; Bauer et al. 2011) and regional climate models (Feldmann et al. 2008; Kendon et al. 2012; Warrach-Sagi et al. 2013) confirmed the improvement in model performance.
A variety of high-resolution studies of land surface–vegetation–atmosphere feedback processes have already been performed. Dalu et al. (1996) found that the daytime convective boundary layer (CBL) starts to be affected by landscape heterogeneity on the scale of 1–10 km. Changes in CBL properties, clouds, and precipitation over this domain as well as mesoscale circulations determine the modification of weather and regional climate (see, e.g., Mahfouf et al. 1987; Pielke et al. 1991; Avissar and Liu 1996; Pielke et al. 2007). Feedback mechanisms initiated by land-use changes are nonlinear as a result of simultaneous changes of surface albedo and evapotranspiration. The partitioning of surface fluxes changed the boundary layer depth and humidity, which eventually led to the formation of clouds and precipitation.
Observational studies confirmed these subtle relationships, for example, due to heterogeneity of soil moisture, land use, and vegetation properties (Bougeault et al. 1991; Lyons et al. 1993; Rabin and Martin 1996). Cutrim et al. (1995) found indications of changes in cloud cover due to deforestation and Spracklen et al. (2012) found a strong relationship between land cover and precipitation in tropical regions. A step further was made in Weaver and Avissar (2001), where changes in land surface properties by human influence (agriculture) were analyzed and an influence on the atmospheric boundary layer (ABL) and cloud development was detected.
Although the effects of tropical deforestation have been the subject of various studies (e.g., Henderson-Sellers et al. 1993; Polcher and Laval 1994; Cutrim et al. 1995; Spracklen et al. 2012), similar research in arid regions is not so prevalent. For instance, Perlin and Alpert (2001) investigated the impact of past and simulated land-use changes in Israel on convection and found a dependence on land surface conditions. More studies in other regions dependent on different meteorological forcing conditions are needed. Particularly, we are not aware of an afforestation study focusing on the understanding of land surface–atmosphere feedback specifically in dry coastal regions for optimizing the weather and climate modification. This is the research goal we are addressing with this work.
The scarcity of previous studies is likely due to the limited feasibility of the reforestation of large arid areas (e.g., irrigation with low environmental impact). Profound knowledge in the cultivation of suitable plants such as Jatropha curcas was also lacking. Major strides in agricultural and crop engineering within the last decade (Li et al. 2010) mean that it is now reasonable to simulate the impact of planted areas on weather and climate. This includes the operation and modification of high-resolution land surface–vegetation–atmospheric models to simulate the irrigation of plants.
Based on these considerations, we prefer the small-scale to large-scale approach for studying the impact of land surface modifications over large domains; the first priority was set on temporally and spatially highly resolved quantitative simulations of the water cycle with a convection-permitting model, which demonstrated high skill with respect to the simulation of precipitation (Weisman et al. 2008; Schwitalla et al. 2011). This permits the verification of model results using small-scale plantations and the subsequent increase of their domains while maintaining high confidence in the simulations. In the model system, the size of the plantation can be enlarged to the entire domain of unused land in desert regions where irrigation is possible by desalination plants or waste water usage. In this initial study, we are focusing on a size of 10 000 km2 for simulating carbon farming and afforestation. Our results can be considered to be baseline information concerning the potential of large-scale plantations for CEA.
b. Climate conditions in Oman and the Sonora
Two desert regions located in the Northern Hemispheric subtropics were selected, the Oman Desert at about 20°N along the eastern coast of the Arabian Peninsula and the Sonora Desert at about 32°N between the northern coast of the Gulf of California and the U.S. border. Because of their location in the subtropical belt, the insolation in the two regions is very similar in summer. In winter, the differences in latitude lead to stronger insolation in Oman. Although both locations are deserts in the Northern Hemispheric subtropics, the climatological conditions in the two regions clearly differ (Warner 2004).
In the Oman Desert, relatively flat terrain dominates with elevations less than 200 m. The temperatures are high throughout the year. The climate diagram at the Masirah airport station (20.65°N, 58.88°E) on an island at the eastern coast of Oman shows temperatures between 25° and 30°C throughout the year with the maximum in summer. The annual precipitation sum is 35 mm without a significant annual cycle. In winter, along the coast of Oman, cooler northeasterly winds blow from the Asian continent over the Arabian Sea, merging with a northeasterly flow from the land (Fig. 1, top-right panel, for December 2007). Because of the larger friction over the land surface, the wind direction is deflected toward the continent whereas the winds remain parallel to the coastline over the ocean. This causes low-level divergence and subsidence along the coast but a slight convergence zone is formed over land. The 2-m mixing ratio is ≈14 g kg−1 in this region. In summer, the wind direction changes and strong southwesterly winds are prevalent along the coast of Oman, where they collide with a stronger northeasterly flow from the land (Fig. 1, top panel, for July 2007). Although the divergence effect along the coast is also present in summer, the convergence zone over land becomes stronger as a result of the development of a heat low over the southeastern part of the Arabian Peninsula. Therefore, the humidity along the coast increases to values of ≈18 g kg−1, leading to more favorable conditions for convection initiation during summer.
(top) The 10-m wind field and 2-m mixing ratio during (left) July and (right) December 2007 over the Arabian Peninsula. The location of the plantation is indicated by the red circle. (bottom) Corresponding fields for (left) July and (right) December for the Gulf of California. The location of the plantation is also shown. The atmospheric fields were extracted from operational ECMWF analyses.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
(top) The 10-m wind field and 2-m mixing ratio during (left) July and (right) December 2007 over the Arabian Peninsula. The location of the plantation is indicated by the red circle. (bottom) Corresponding fields for (left) July and (right) December for the Gulf of California. The location of the plantation is also shown. The atmospheric fields were extracted from operational ECMWF analyses.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
(top) The 10-m wind field and 2-m mixing ratio during (left) July and (right) December 2007 over the Arabian Peninsula. The location of the plantation is indicated by the red circle. (bottom) Corresponding fields for (left) July and (right) December for the Gulf of California. The location of the plantation is also shown. The atmospheric fields were extracted from operational ECMWF analyses.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Most of the year, the Sonora is dominated by the North Pacific subtropical high to the west (Fig. 1, bottom). It blocks low pressure systems from reaching the west coast and the large-scale subsidence suppresses the development of local convection. The high pressure system is shifted with season. In summer it is strong and located at about 38°N. During winter it is weaker and centered off the coast of lower California at around 20°–25°N. A mountain range along lower California, reaching elevations up to 2000 m, stabilizes the warm and dry conditions throughout the year. Winter storms from the Pacific are generally weakened. The rain shadow effect increases the stability of the downwind atmosphere and reduces the water vapor available for the development of precipitation (Warner 2004). In winter, northerly winds dominate, transporting dry and much colder air from the elevated deserts in the United States into the region and suppressing convection and precipitation. In summer, strong heating occurs in the Sonora Desert, causing the development of a weak heat low. As a consequence, the circulation over the Gulf of California changes and with southeasterly winds, moist air is advected along the Gulf of California into the region of the Sonora Desert. This transport is part of the North American monsoon system transporting moisture from the tropical Pacific and the Gulf of Mexico into the region. Because of the cool northerlies in winter and the strong heating in summer, the annual cycles of temperature, moisture, and precipitation are stronger in the Sonora than in Oman. For instance, at the Puerto Peñasco station at the northern tip of the Gulf of California (31.35°N, 113.52°W), maxima in summer reach 35°C and ≈18 g kg−1. In contrast, minima in winter of 15°C and ≈5 g kg−1 were observed. With 76 mm, the annual amount of precipitation received is twice as much as in Oman with a minimum in late spring.
For our study, the year 2007 was chosen to utilize the most recent operational analysis data available from the European Centre for Medium-Range Weather Forecasts (ECMWF) at the time. The representativeness of conditions in 2007 against a 30-yr mean was compared in terms of mean monthly 2-m air temperatures, using data from the closest weather stations to the plantations, Masirah in Oman and Puerto Peñasco in the Sonora (not shown). Monthly temperatures in 2007 varied only slightly from the long-term mean in both the annual pattern and magnitudes. Deviations throughout the year are generally 0.5°C or less in Masirah. At Puerto Peñasco, 2007 temperatures were similar in summertime (slightly warmer by 0.2°–1°C) but substantially cooler in winter than the mean (up to 3°C). This could be a result of the strong La Niña effect present in 2007 or just due to interannual variability. As will become apparent, the emphasis of this paper centers on the summer period (June–August: JJA), in which case the representativeness of the year 2007 was satisfactory in terms of temperatures for both central Oman and the Sonoran Desert.
3. Modeling strategy
The simulations were performed with an advanced, coupled land surface–vegetation–atmospheric model (Weather Research and Forecasting–Noah land surface model: WRF–Noah) on the convection-permitting scale. Simulations were run for a full year (2007) in order to assess feedback processes with high statistical confidence during all seasons. In particular, we investigated the modification of surface variables such as flux partitioning by the EBC scheme, temperature, and humidity. Furthermore, the boundary layer depth, the cloud coverage, the rain rate, and the formation of dew were analyzed. For improving process understanding, we present a case study of a convection initiation event in 4D and the chain of events leading to the formation of clouds and precipitation. Some seasonally averaged results of these simulations were already presented in Becker et al. (2013). Here, we are focusing on a detailed analysis of feedback processes in the land surface–vegetation–atmosphere systems and a case study of convection initiation.
a. Setup and adaptation of the coupled model system
Version 3.1 of the WRF model coupled to the Noah land surface model (LSM) was employed. Recent studies demonstrated a good level of performance for this model system with respect to the simulation of soil moisture (Greve et al. 2013).
To minimize systematic errors due to large-scale conditions, WRF–Noah was driven by operational analyses of the ECMWF. The nested domain was chosen to be a few tens of kilometers larger than the planted region in the center of the domain. At the boundaries, one-way nesting was applied because we assumed that the scale of the planted region was still too small to induce large-scale feedback processes to the exterior domain. How large the inner domain must become before this assumption deteriorates is left to future studies.
It was possible to avoid additional nests by application of ECMWF data on a 12.5-km grid, as good performance of the nested WRF with a grid increment ratio of 3 between the boundaries and the inner domain was demonstrated. The sizes of the plantations were 10 000 km2 in both cases. Although the land area used was the same in both regions, the plantation geometries were different mainly because of land-use types and properties in their respective locations. In Oman, it was possible to find a nearly square area on arid land, and in the Sonora, a more flattened rectangular shape was necessary, limited by the U.S.–Mexico border between and surrounding land use.
Based on previous and ongoing work on the verification of various combinations of parameterizations (e.g., Schwitalla et al. 2011; Warrach-Sagi et al. 2013; Greve et al. 2013; Vautard et al. 2013; Milovac et al. 2013), the following combination of parameterizations was chosen: the shortwave radiation by Dudhia (1989), the longwave radiation with the Rapid Radiative Transfer Model (RRTM) model by Mlawer et al. (1997), the Yonsei University (YSU) boundary layer turbulence scheme (Hong et al. 2006), and the Morrison cloud microphysics (Morrison et al. 2009). The Noah LSM (Chen and Dudhia 2001) contains a Penman-based energy balance approach including a stability-dependent aerodynamic resistance (Mahrt and Ek 1984; Chen et al. 1996). The WRF model configuration and its parameterizations are presented in Tables 1 and 2. We performed a 1-yr downscaling experiment driven by ECMWF analyses during 2007. We applied this setup to study the changes in the statistics of land surface and atmospheric variables by the plantations by performing a CONTROL run without changes to the land surface properties and an IMPACT run with implementation of the plantations.
The configuration of the WRF model.
Parameterizations chosen in the WRF model.
b. Incorporation of the jatropha vegetation properties and irrigation
A key topic of this study was the adaptation of Noah to the regions of interest as well as to the modification of land surface properties by the plantations. For this purpose, we initialized the model with the advanced 20-category vegetation–land-use International Geosphere-Biosphere Programme (IGBP) Moderate Resolution Imaging Spectroradiometer (MODIS) dataset (Friedl et al. 2002). A modification of the vegetation parameters of Jatropha curcas and a change in the handling of irrigation were implemented in the Noah LSM in order to achieve most realistic results.
From the 20 categories in the IGBP MODIS data, the vegetation type evergreen broadleaf forest comes closest to Jatropha curcas. We optimized the parameter set for this vegetation type so that it was most consistent with recent published vegetation properties of jatropha. The maximum leaf-area index was reduced from 6.48 to 3.2 and the minimum leaf-area index was increased from 3.08 to 3.2 (Holl et al. 2007). The standard minimum stomata resistance of 150 s m−1 was set to 132 s m−1 (Hänel and Löpmeier 1997). The main difference between the vegetation and the desert concerns the decrease of the albedo from 0.38 to 0.12, but the increase of the roughness length from 0.01 to 0.5 m can also play a role in the modification of the surface wind field. The albedo may be subject to change too if suitable measurements become available, but these results had not yet been published as of the time of our simulations. The resulting jatropha parameters are presented in Table 3. Remaining systematic deviations were not considered critical for this study, as these were expected to affect the resulting impact of the vegetation cover on atmospheric variables to second order.
Vegetation parameters for evergreen broadleaf forest and its optimization for jatropha. Land-use type 2 is for evergreen broadleaf forest, with the standard values shown in parentheses; modifications are marked in boldface. References either confirm the present or the changed values. In comparison, land-use type 16 (barren or sparsely vegetated) is also shown.
Another requirement was a reasonable consideration of irrigation. For this purpose, any water stress of the plant should be avoided by setting the soil moisture at least to the wilting point. Within the land surface model code, we constrained the volumetric soil moisture content so that it did not fall below 0.08, which was higher than the wilting point of loam (0.06) or sandy loam (0.047), the two dominant soil textures in the regions of interest. The remaining effect of irrigation can be estimated as follows. Based on real field results presented in Becker et al. (2013), the irrigation amount per year for Jatropha curcas is rather low (≈100 mm yr−1). This translates into ≈0.27 mm day−1 = 6.25 × 10−6 L m−2 s−1 ≈ 16 W m−2 maximum available latent heat flux by evapotranspiration during daytime, if the soil is dry and no precipitation is present. This was mainly the case in summer. Consequently, the influence of area-averaged irrigation on the surface latent heat flux can be neglected, and this is still the case, if even considerably higher irrigation amounts are considered.
Figure 2 present the soil textures in Oman and the Sonora. The locations of the plantations at the coasts in Oman and Mexico are easily detected. Large-scale maps of the locations of the plantations and the changes in albedo are presented in Becker et al. (2013). The significant changes of the albedos are also indicated in the captions. For soil texture, Food and Agriculture Organization of the United Nations (UN/FAO) data were used (IUSS 2007) and are included in the preprocessing package of WRF. In Oman, in the plantation region, the soil is sandy loam (ISLTYP = 3) and loam (ISLTYP = 6) (Fig. 2, top), whereas in the Sonora it is sandy clay loam (ISLTYP = 7) (Fig. 2, bottom). However, it is certainly suspicious that the soil texture changes along the border between Mexico and the United States. We did not make an attempt to correct this because, on the one hand, data with higher quality are missing and, on the other hand, the effect is a small and due to the low and constant soil moisture results in a relatively constant bias of the soil heat flux.
The 2D plots of the soil texture in the (left) Oman and (right) Sonora modeling domains. The shape of the simulated plantation is also indicated by the green line. Because of the land-use change, the albedo decreases from ≈0.38 in Oman and ≈0.23 in the Sonora to ≈0.13 over the plantation.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
The 2D plots of the soil texture in the (left) Oman and (right) Sonora modeling domains. The shape of the simulated plantation is also indicated by the green line. Because of the land-use change, the albedo decreases from ≈0.38 in Oman and ≈0.23 in the Sonora to ≈0.13 over the plantation.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
The 2D plots of the soil texture in the (left) Oman and (right) Sonora modeling domains. The shape of the simulated plantation is also indicated by the green line. Because of the land-use change, the albedo decreases from ≈0.38 in Oman and ≈0.23 in the Sonora to ≈0.13 over the plantation.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
c. Performance of model runs
To study the impact of the modification of the vegetation properties in the domains of interest with high statistical certainty, four simulations were performed for the entire year 2007: Oman CONTROL (no changes to the vegetation properties), Oman IMPACT (modification of land surface properties by a jatropha plantation in selected regions) as well as Sonora CONTROL and Sonora IMPACT. The Oman CONTROL and IMPACT runs were performed in a region from 18.96° to 21.63°N and 55.88° to 58.72°E with 79 × 79 grid points at 4 km horizontal grid increments. The Sonora runs were performed covering 30.19°–32.73°N and 111.92°–114.89°W using 74 × 74 grid points at 4-km increments as well (see Table 1). To the best of our knowledge, these are the longest duration convection permitting simulations to have been performed in these regions to date.
Eight variables were selected in order to detect feedback processes in the land surface–vegetation–atmosphere system: 2-m atmospheric temperature, 2-m-atmospheric specific humidity, surface sensible heat flux, surface latent heat flux, ground heat flux, ABL depth, cloud coverage, and accumulated precipitation. For each simulation, seasonal averages and their standard deviations were determined for all these variables. Furthermore, for these eight variables, mean diurnal cycles and their standard deviations were produced. The diurnal cycles were only averaged in the region where the vegetation properties were changed. For the presentation of the surface energy balance, the micrometeorological convention was used that fluxes directed away from the surface are positive and the transport of energy toward the surface is negative.
4. Results
a. Oman site statistics during summer and winter
The difference in albedos between CONTROL and IMPACT led to substantial changes in surface fluxes and variables as well as CBL development over the plantation. Figure 3 presents the diurnal cycles of the fluxes contributing to the EBC for summer. In the top panel of Fig. 3, the surface energy balance over the desert was simulated. During daytime, which is the time period with positive solar incoming radiation, the net radiation, which reached ≈−450 W m−2 around local noon, was balanced by the surface heat flux and the ground heat flux. During nighttime, the positive net radiation loss of 50–100 W m−2 was mainly balanced by the ground heat flux.
Diurnal cycles of the surface energy balance in Oman: (top) CONTROL, or desert conditions; (bottom) IMPACT, or conditions over the plantation.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Diurnal cycles of the surface energy balance in Oman: (top) CONTROL, or desert conditions; (bottom) IMPACT, or conditions over the plantation.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Diurnal cycles of the surface energy balance in Oman: (top) CONTROL, or desert conditions; (bottom) IMPACT, or conditions over the plantation.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Over the plantation, the partitioning of the fluxes changed considerably. During daytime, the net radiation decreased to ≈−650 W m−2 at noon. This resulted in an increase in the sensible heat flux to 600 W m−2 at local noon whereas the ground heat flux and the latent heat flux remained <50 W m−2, respectively. For the ground heat flux this was due to the dense vegetation cover, for the latent heat flux this was due to the low soil moisture and a reduction of transpiration at high temperatures. During nighttime, the ground heat flux remained low so that the positive net radiation was mainly balanced by a negative sensible heat flux. This led to a stronger cooling of the vegetated surface than over the desert.
The fluxes contributing to the EBC during wintertime are shown in Fig. 4. Over the desert (Fig. 4, top), during daytime, less net radiation was available (≈−350 W m−2) so that the sensible heat flux and the ground heat flux decreased, too. The relative contribution of the ground heat flux balancing net radiation is higher than during summertime. During nighttime, the net radiation was similar to that during summer. It was balanced by a combination of the ground heat flux and the sensible heat flux with a slightly higher contribution of the latter than during summer.
As in Fig. 3, but for winter.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
As in Fig. 3, but for winter.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
As in Fig. 3, but for winter.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Over the plantation (Fig. 4, bottom), during daytime, the net radiation decreased to ≈−550 W m−2 at noon so that an albedo effect still existed during this season. But now the plants were permitted to transpire more water so that the latent heat flux became an important contributor to the surface energy balance whereas the ground heat flux remained <50 W m−2. During nighttime, the net radiation was smaller than over the desert, indicating a low temperature over the plantation. Now, the ground heat flux, the sensible heat flux, and even the latent heat flux were contributing to the balance. The latter led to the formation of dew. Negative values of the latent heat flux were found in spring and fall, too (not shown). This effect may be considered to use dew formation for reducing irrigation amounts.
The resulting impacts on 2-m temperature and 2-m mixing ratio in Oman are presented in Fig. 5. Both in winter and summer, the amplitude of the diurnal cycle of 2-m temperature was enhanced. Temperatures increased over the vegetation during daytime by 1–2 K and decreased by 2–4 K during nighttime in summer and winter, respectively. This response was nonlinear and resulted in an effective reduction of the mean temperature over the plantation. This temperature decrease was relatively small in summer (about 0.3 K) but substantial in winter (about 1.8 K). This is an interesting effect, which may be considered for climate change mitigation.
Diurnal cycles of 2-m atmospheric (left) temperature and (right) specific humidity during (top) summer and (bottom) winter in Oman.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Diurnal cycles of 2-m atmospheric (left) temperature and (right) specific humidity during (top) summer and (bottom) winter in Oman.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Diurnal cycles of 2-m atmospheric (left) temperature and (right) specific humidity during (top) summer and (bottom) winter in Oman.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
The reduction in humidity between summer and winter confirmed the large-scale analyses presented in Fig. 1 and discussed in section 2b. However, the response of 2-m specific humidity to the plantation was rather complex. In summer, during daytime, the humidity was reduced by about 1 g kg−1. This may appear counterintuitive but it is reasonable as long as the latent heat flux is small and the sensible heat flux large leading to strong vertical mixing of surface humidity in a deeper CBL. In winter, during daytime, stronger evapotranspiration occurred, leading to a slight increase in humidity, whereas during nighttime, a reduction of humidity was found and was likely due to a negative latent heat flux.
The results confirm the significant and nonlinear modification of the regional water and energy cycle by the plantation, requiring high-resolution, high quality 3D simulations. Also, dynamics in the region was influenced by the modification of the land surface properties. Increased surface roughness and sensible heat led to more vigorous vertical mixing and the development of a higher ABL top. This is demonstrated in Fig. 6, which shows the spatially resolved increase of the IMPACT ABL depth over Oman during summer. The mean ABL depth over the plantation was higher by several 100 m in comparison to the surrounding area. The ABL depth was subject to a strong diurnal cycle with an increase from 2000 m over the desert to 2700 m over the plantation around local noon (not shown) following the diurnal cycle of the sensible heat flux. The increase of the ABL depth over the plantation was spatially not homogeneous but was due to the development of an internal thermal CBL. Therefore, the increase of the daytime CBL depth was lower upstream and higher downstream of the monsoon flow, respectively (Fig. 6).
Mean IMPACT ABL depth (m MSL) over Oman in summer.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Mean IMPACT ABL depth (m MSL) over Oman in summer.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Mean IMPACT ABL depth (m MSL) over Oman in summer.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
The increased CBL top resulted in a higher likelihood to reach the lifting condensation level and in a higher convective instability, respectively. These effects are studied in detail in section 4c. Indeed, a higher frequency of clouds was simulated (not shown). Whereas the region was essentially cloud free in summer, the cloud frequency increased to about 1% over the plantation. This corresponded to at least one cloud occurrence in each cell column over the domain during the summer.
Of particular interest was the potential modification of precipitation over the plantation. Figure 7 shows the IMPACT and CONTROL precipitation amounts accumulated during summer. An increase in precipitation over the plantation was simulated. This effect only occurred in summer and was connected with the enhancement of the airmass vertical instability and with a modification of the surface flow. The amount of the simulated change of precipitation is expected to be dependent on the size of the plantation. Using an area of 100 km × 100 km, the precipitation enhancement averaged over the domain was ≈10 mm in summer. This corresponded to a doubling of the precipitation amount in comparison to the CONTROL run. The precipitation enhancement was very complex and heterogeneous (Fig. 7). As expected in the real world, this was due to a series of single convective events, which are identified in section 4c.
Summer precipitation for (top) IMPACT and (bottom) CONTROL. The labels show the precipitation amounts in millimeters per season.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Summer precipitation for (top) IMPACT and (bottom) CONTROL. The labels show the precipitation amounts in millimeters per season.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Summer precipitation for (top) IMPACT and (bottom) CONTROL. The labels show the precipitation amounts in millimeters per season.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
b. Sonora site statistics during summer and winter in comparison with Oman
Figures 8–10 present the resulting diurnal cycles of the fluxes contributing to the EBC for summer and winter, respectively, as well as the diurnal cycles of 2-m temperature and 2-m specific humidity. As expected from the analyses in section 2b, an EBC closure similar to that in Oman could be expected.
Diurnal cycles of the surface energy balance in the Sonora: (top) CONTROL, or desert conditions; (bottom) IMPACT, or conditions over the plantation.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Diurnal cycles of the surface energy balance in the Sonora: (top) CONTROL, or desert conditions; (bottom) IMPACT, or conditions over the plantation.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Diurnal cycles of the surface energy balance in the Sonora: (top) CONTROL, or desert conditions; (bottom) IMPACT, or conditions over the plantation.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
As in Fig. 8, but for winter.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
As in Fig. 8, but for winter.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
As in Fig. 8, but for winter.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Diurnal cycles of 2-m atmospheric (left) temperature and (right) specific humidity during (top) summer and (bottom) winter in the Sonora.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Diurnal cycles of 2-m atmospheric (left) temperature and (right) specific humidity during (top) summer and (bottom) winter in the Sonora.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Diurnal cycles of 2-m atmospheric (left) temperature and (right) specific humidity during (top) summer and (bottom) winter in the Sonora.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Indeed, during summertime over the desert and during daytime, the net radiation reached ≈−450 W m−2, balanced by the sensible heat flux and the ground heat flux. The slope of the ground heat flux during the morning transition was higher than in Oman, which is likely due to the slightly different soil texture and moisture. During nighttime, a similar net radiation (50–100 W m−2) mainly balanced by the ground heat flux was found.
Over the plantation during daytime, the net radiation decreased to ≈−630 W m−2 at noon, resulting in an increase in the sensible heat flux (550 W m−2) while the latent heat flux and the ground heat flux remained low (<50 W m−2). The contribution of the ground heat flux was slightly higher than in Oman, which explains the difference in the sensible heat flux at noon. During nighttime, the net radiative loss was ≈50 W m−2, balanced by the sensible heat flux and the ground heat flux with nearly similar contributions.
During winter over the desert (see section 2b and Fig. 9, top) and during daytime, significantly less net radiation was available than in Oman (≈−280 W m−2 around noon). The balance was maintained by the ground heat flux and the sensible heat flux with similar contributions. Again the slope of the ground heat flux during the morning transition was higher than in Oman leading to a phase delay in the diurnal cycles of the ground heat flux and the sensible heat flux. During nighttime, the net radiative loss was 50–100 W m−2, mainly balanced by the ground heat flux similar to that in Oman.
Over the plantation (Fig. 9, bottom) during daytime, the albedo effect was still present, leading to a decrease in the net radiation to ≈−400 W m−2. In a relative sense, a similar pattern of behavior for the ground heat flux, the latent heat flux, and the sensible heat flux was found as in Oman. The latent heat flux started to play a substantial part in the surface energy balance reaching ≈130 W m−2 at local noon. During nighttime, the net radiative loss and the sensible heat flux were slightly higher than in Oman but otherwise the balance turned out to be similar.
The diurnal cycles of 2-m atmospheric temperature and 2-m atmospheric specific humidity in the Sonora are presented in Fig. 10. During summertime over the desert, a qualitatively similar pattern of behavior for temperature compared with Oman was simulated except that the absolute temperatures were higher in Oman by about 2 K. During winter, the differences in temperature between Sonora and Oman were higher and amounted to approximately 9 K, mainly driven by different surface heating. Over the plantation, the amplitudes of the diurnal cycle of temperature were intensified in summer and in winter, respectively. In summer during nighttime over the plantation, a stronger cooling was found in the Sonora than in Oman. This may be due to a slightly more negative ground heat flux in the Sonora. In the Sonora over the plantation, mean temperature reductions of 1.25 K in summer and 1.57 K in winter were simulated. Also in summer, the increase of evapotranspiration was not sufficient to enlarge the 2-m atmospheric specific humidity throughout the day. In winter, a slight increase in humidity was simulated but a small reduction was found during the nighttime as well. Both in summer and winter, the humidity was significantly less in the Sonora by ≈2–3 g kg−1 and even 5 g kg−1, respectively, in contrast to Oman, as expected from the analyses presented in section 2b.
Figure 11 presents the mean summer ABL depth of the IMPACT run in the Sonora. The ABL depth increased by about 210–1200 m, mainly over the central part of the plantation (see Becker et al. 2013). Upstream of the plantation and at the southern edges of the plantation, a slight reduction of ABL depth was simulated due to the development of a daytime thermal internal CBL. In this region, the mean ABL depth increased from 1800 m over the desert to 2600 m over the plantation at local noon. This mechanism is very similar to that in Oman. Again, the ABL depth was not constant but was maximal mainly in the northwestern and southeastern corners of the plantation caused by the mean wind direction from the southeast. Downstream, the ABL depth remained higher than in CONTROL for about 20 km. Here, the internal ABL was mixing in a complex way with the less convective downstream ABL. This effect was different between Oman and the Sonora and depended in a complex way on the vertical stability of the air masses over and downstream of the plantations as well as on the vertical stability and subsidence above the ABL.
Mean IMPACT ABL depth (m MSL) in summer over the Sonora.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Mean IMPACT ABL depth (m MSL) in summer over the Sonora.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Mean IMPACT ABL depth (m MSL) in summer over the Sonora.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Figure 12 presents the IMPACT and CONTROL precipitation amounts accumulated during summer in the Sonora. A significant increase in precipitation over the plantation was simulated, reaching 160 mm at specific locations. Thus, the precipitation increase was stronger than in Oman. The heterogeneous structures indicate that the increase in precipitation was due to several convection initiation events. But there was also a reduction in precipitation downstream of the plantation. This demonstrates that in the Sonora the enhancement of precipitation was not only due to convection initiation of previously nonexistent precipitation events but also due to the enhancement, modification, and displacement of precipitation, which was already present in CONTROL.
Sonora summertime precipitation: (top) IMPACT and (bottom) CONTROL precipitation amounts. The labels show the precipitation amounts in millimeters per season.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Sonora summertime precipitation: (top) IMPACT and (bottom) CONTROL precipitation amounts. The labels show the precipitation amounts in millimeters per season.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Sonora summertime precipitation: (top) IMPACT and (bottom) CONTROL precipitation amounts. The labels show the precipitation amounts in millimeters per season.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
c. Study of convection initiation: Oman, 7 June 2007
It is important to understand the combined effect of surface forcing processes leading to convective events. This gives insights into whether convection initiation in the model is simulated in a realistic and reasonable way. Particularly, trigger mechanisms in the model leading to deep convection, such as the formation of convergence zones, need to be analyzed with respect to process understanding. We stress that a modification of the vertical stability of the air mass over the plantation is a necessary but not sufficient condition for convection with significant amounts of precipitation. Only if a strong and temporally and spatially persistent trigger mechanism is present can deep and coherent vertical convection be initiated and maintained. We study whether the land surface heterogeneity of the plantation inserted in the desert can be the driver of these processes by means of 3D visualization of several land surface and atmospheric variables.
All simulations for summer were analyzed from day to day in both regions. Only in this season was a significant precipitation enhancement observed. For each of the 92 days, the precipitation amount was calculated and a comparison between IMPACT and CONTROL was performed. Only a few cases contributed to the precipitation enhancement. Tables 4 and 5 summarize the most important events. These cases are ordered according to the strength and the evidence of whether land surface changes triggered the event. This is particularly clear when convection initiation took place in the IMPACT run only. In the Oman simulations, it was found that convection initiation took mainly place either at the northeastern or the southwestern corner of the plantation. Depending on the wind shear between the CBL and the free troposphere, a convective system developed, which propagated southeast or north into the plantation, respectively. In the Sonora, the convection was triggered by or amplified an existing event either along the northern border of the plantation or in the southeast of the plantation.
Important precipitation events in Oman demonstrating that the precipitation enhancement is due to several isolated convection initiation events in the IMPACT runs.
Important precipitation events in the Sonora showing that the precipitation enhancement is due to convection initiation and amplification during several events in the IMPACT runs.
In the following, we select the Oman case 1 from Table 4 for studying convection initiation as well as cloud and precipitation development on 7 June 2007. We focus on 7 June 2007 because on this day localized convection initiation took place over the plantation in IMPACT while almost no precipitation was found in CONTROL.
The vertical stability can be studied by calculating 2D fields of the convective available potential energy (CAPE) and of the convective inhibition (CIN). These results are presented in Fig. 13 for 0700 UTC shortly before convection initiation. Substantial differences in the CAPE fields were simulated, including a high spatial variability in the region of the plantation. In the IMPACT run, upstream and in the southern part of the plantation, the CAPE was higher by about 200 J kg−1, reaching values of 1750 J kg−1. Downstream and in the northern part of the planted domain, the CAPE was lower in IMPACT by about 200–400 J kg−1. This spatial structure was strongly related to the development of the thermal internal CBL over the plantation. In contrast, the CIN field did not exhibit such a strong spatial dependence. But it was significantly reduced in IMPACT by ≈40 J kg−1 so that the CIN was virtually eliminated by the change of the thermodynamical structure of the CBL. This effect substantially increased the likelihood of convection initiation in the IMPACT run.
(left) CAPE and (right) CIN fields in Oman at 0700 UTC 7 Jun 2007, shortly before convection initiation for (top) CONTROL and (bottom) IMPACT.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
(left) CAPE and (right) CIN fields in Oman at 0700 UTC 7 Jun 2007, shortly before convection initiation for (top) CONTROL and (bottom) IMPACT.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
(left) CAPE and (right) CIN fields in Oman at 0700 UTC 7 Jun 2007, shortly before convection initiation for (top) CONTROL and (bottom) IMPACT.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
For the development of a convective precipitating system, a trigger mechanism such as a convergence zone must be present so that the resulting vertical lifting delivers a coherent updraft overcoming CIN and reaching the lifting condensation level. Indeed, a convergence zone developed, which was related to the changes in the land surface properties. The modified surface energy balance led to a substantial increase in the surface temperature, as discussed above. Figure 14 (top) presents the differences in T2 at 0700 UTC. The temperature in IMPACT was about 2°–3°C higher, particularly in the northwestern corner of the plantation.
(top) Difference in 2-m atmospheric temperature between IMPACT and CONTROL in Oman at 0700 UTC 7 Jun 2007, shortly before convection initiation was initiated. (bottom) Corresponding sea level pressure in (left) CONTROL and (right) IMPACT.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
(top) Difference in 2-m atmospheric temperature between IMPACT and CONTROL in Oman at 0700 UTC 7 Jun 2007, shortly before convection initiation was initiated. (bottom) Corresponding sea level pressure in (left) CONTROL and (right) IMPACT.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
(top) Difference in 2-m atmospheric temperature between IMPACT and CONTROL in Oman at 0700 UTC 7 Jun 2007, shortly before convection initiation was initiated. (bottom) Corresponding sea level pressure in (left) CONTROL and (right) IMPACT.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
The impact on the sea level pressure field is demonstrated in Fig. 14 (bottom). The CONTROL run showed a typical pressure gradient between the ocean and the land surface (Fig. 14, bottom left). Some fine structures were likely due to differences in soil properties. The increased temperature over the plantation produced a heat low with a significant pressure reduction over the plantation of up to 1 hPa (Fig. 14, bottom right).
The pressure and surface drag modifications had important consequences. Figure 15 presents the structure of the low-level temperature and wind fields at 0700 UTC just before deep convection was triggered in IMPACT while convection initiation was absent in CONTROL. The 2-m temperature structures over land in CONTROL such as in the northwestern corner of the model region are likely due to different soil properties (Fig. 2, left). In both panels in Fig. 15, the wind fields outside the plantation region show the summer monsoon flow from the southwest, which was nearly parallel to the coastline in agreement with the pressure field shown in Fig. 14. This was beneficial for the advection of relatively moist air in the region of the plantation. Just in the region of the plantation, Fig. 15 shows a striking difference in the low-level wind field. Because of different surface pressure fields but also because of different surface drag, a strong flow deflection was caused in IMPACT. The pressure gradient and the friction were bending the surface flow significantly toward the plantation, leading to a convergence zone at its lee side. This is the coherent trigger mechanism we were looking for so that deep convection could be initiated and maintained, overcoming the remaining CIN in the lower troposphere.
The 300-m flow and 2-m temperature fields in the (top) CONTROL and (bottom) IMPACT runs at 0700 UTC 7 Jun 2007. As expected, the wind speed is higher over the sea surface than over land.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
The 300-m flow and 2-m temperature fields in the (top) CONTROL and (bottom) IMPACT runs at 0700 UTC 7 Jun 2007. As expected, the wind speed is higher over the sea surface than over land.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
The 300-m flow and 2-m temperature fields in the (top) CONTROL and (bottom) IMPACT runs at 0700 UTC 7 Jun 2007. As expected, the wind speed is higher over the sea surface than over land.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Figure 16 presents a vertical cross section of the 3D rendered humidity field shortly before convection initiation took place. The 2-m temperature field and the wind barbs at 300 m are also shown. The humidity mixing ratio in the CBL was significant and amounted up to 15 g kg−1. An internal thermal CBL with an increasing top along the low-level wind direction and wake effects downstream of the plantation is visible. The exact form was determined by the wind direction and speed as well as by the vertical stability. The maximum depth of the CBL was reached in the region of the convergence zone.
Cross section through 3D-rendered humidity field over the 2-m atmospheric temperature at 0700 UTC 7 Jun 2007. The wind barbs at 300 m MSL are also shown.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Cross section through 3D-rendered humidity field over the 2-m atmospheric temperature at 0700 UTC 7 Jun 2007. The wind barbs at 300 m MSL are also shown.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Cross section through 3D-rendered humidity field over the 2-m atmospheric temperature at 0700 UTC 7 Jun 2007. The wind barbs at 300 m MSL are also shown.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Figure 17 presents the evolution of the convective system at 0800 UTC (top panel) and 1000 UTC (bottom panel). The 2-m temperature field overlaid by the 300-m wind field as in Fig. 15 are shown but in addition isosurfaces for vertical wind, cloud liquid water, ice clouds, and precipitation are presented in 3D for studying the interaction of convection initiation as well as cloud and precipitation processes.
(top) Low-level wind and 2-m temperature fields as well as red and gray isosurfaces of vertical wind speed and cloud liquid water, respectively, at 0800 UTC when convection was initiated. The isosurface values are 0.6 m s−1 and 0.074 g kg−1, respectively. (bottom) Well-developed convective system at 1000 UTC including green isosurfaces for ice clouds (1.6 μg kg−1) and blue isosurfaces for rainwater (0.1 g kg−1). The color scales indicate the wind speed of the flow field for the top panel and the 2-m atmospheric temperatures for the bottom panel.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
(top) Low-level wind and 2-m temperature fields as well as red and gray isosurfaces of vertical wind speed and cloud liquid water, respectively, at 0800 UTC when convection was initiated. The isosurface values are 0.6 m s−1 and 0.074 g kg−1, respectively. (bottom) Well-developed convective system at 1000 UTC including green isosurfaces for ice clouds (1.6 μg kg−1) and blue isosurfaces for rainwater (0.1 g kg−1). The color scales indicate the wind speed of the flow field for the top panel and the 2-m atmospheric temperatures for the bottom panel.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
(top) Low-level wind and 2-m temperature fields as well as red and gray isosurfaces of vertical wind speed and cloud liquid water, respectively, at 0800 UTC when convection was initiated. The isosurface values are 0.6 m s−1 and 0.074 g kg−1, respectively. (bottom) Well-developed convective system at 1000 UTC including green isosurfaces for ice clouds (1.6 μg kg−1) and blue isosurfaces for rainwater (0.1 g kg−1). The color scales indicate the wind speed of the flow field for the top panel and the 2-m atmospheric temperatures for the bottom panel.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
Between 0700 and 0800 UTC, convection initiation took place and evolved in a cloud cluster above the convergence zone. The vertical wind speed below this cluster was about 0.6 m s−1, which was sufficient to overcome any remaining CIN at the CBL top. The cluster was spreading out upstream of the convergence zone as a result of an outflow boundary caused by downdrafts originally produced by convection initiation. The bottom panel in Fig. 17 presents the convective system at 1000 UTC with several isosurfaces for liquid, ice, and rainwater. At this time, the vertical extent of the convective system was reaching the tropopause. Ice phase processes caused a significant amount of rainwater. The rain was mainly falling along an outflow boundary propagating against the mean wind direction. The whole system had a life time of about 6 h and started to decay when reaching the western boundary of the plantation.
The resulting precipitation patterns are presented in Fig. 18. The top panel in Fig. 18 presents the daily integrated precipitation amount of the IMPACT run and the middle panel that of the CONTROL run. Essentially no precipitation was present in the CONTROL run, although some spurious amounts may be present in the northwestern corner of CONTROL. Only in the IMPACT run, convection initiation took place in the northeastern corner of the plantation from the enhanced convergence as outlined above. The main part of the precipitation was induced in the region of the initial convergence zone with ≈20 mm; some further precipitation was produced during the propagation of the convective system to the southwest. Another small cell was triggered west of the plantation before the system was decaying. The bottom panel in Fig. 18 shows the precipitation difference between IMPACT and CONTROL, demonstrating that almost all precipitation was induced in the IMPACT run. Convection initiation was mainly caused by the pressure difference or friction-induced deviation of the surface flow but not by significant changes in the humidity field.
(top) IMPACT and (middle) CONTROL precipitation, as well as (bottom) differences in precipitation between IMPACT and CONTROL on 7 Jun 2007. The color labels show the precipitation amounts in millimeters per day. Note the addition of a negative scale in the bottom panel.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
(top) IMPACT and (middle) CONTROL precipitation, as well as (bottom) differences in precipitation between IMPACT and CONTROL on 7 Jun 2007. The color labels show the precipitation amounts in millimeters per day. Note the addition of a negative scale in the bottom panel.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
(top) IMPACT and (middle) CONTROL precipitation, as well as (bottom) differences in precipitation between IMPACT and CONTROL on 7 Jun 2007. The color labels show the precipitation amounts in millimeters per day. Note the addition of a negative scale in the bottom panel.
Citation: Journal of Applied Meteorology and Climatology 53, 5; 10.1175/JAMC-D-13-0208.1
5. Summary
In this study, the effects of the modification of land surface and atmospheric properties caused by jatropha plantations with an area of about 10 000 km2 in Oman and the Sonora Desert in Mexico were presented and discussed. The influence of plantations in these coastal desert regions was studied by means of a 1-yr downscaling experiment driven by ECMWF boundaries during 2007. The simulations were performed using the WRF–Noah model system without and with a plantation adapted to the land cover properties of jatropha plants and operating on the convection-permitting scale.
The impact of the plantation caused the following effects. The difference between vegetated and unvegetated surface albedo and therefore net surface radiation produced changes of the diurnal cycle of the partitioning of the surface fluxes in the EBC. The modification of the partitioning of surface fluxes depended on the amount of evapotranspiration and the availability of soil moisture. In summer, the contribution of the latent heat flux was small, resulting in strong increases in the sensible heat flux and surface temperature. Interestingly, the mean temperature of the plantations decreased in comparison to the desert situation. This was due to an overcompensation of the increase in temperature during daytime by a strong cooling during nighttime over the plantations. Because of the increase in the sensible heat flux during daytime over the plantations, a deeper internal thermal convective boundary layer developed than over the desert. This increased the likelihood that air parcels reached the lifting condensation level where they induce convection initiation.
During summer both in Oman and in the Sonora, an onset and/or an enhancement of precipitation were simulated, which were caused by several convection initiation events. An Oman case study from 7 June 2007 demonstrated that the precipitation was due to the development of convergence zones at the edges of the plantations. Local heating during daytime over the plantations produced a slight pressure perturbation—a heat low—over the domain. Furthermore, a higher roughness length increased the surface drag. Both effects resulted in a deviation of the surface flow, which eventually led to the development of convergence zones. The shape and the orientation of the domain of the plantation determined the modification of the surface mean flow due to these pressure perturbations, causing local wind and moisture convergence at the edges of the domain and thus increasing the vertical wind speed and moisture transport. The combination of this coherent, mesogamma-scale transport with increased convective instability over the domain resulted in enhanced cloud and precipitation development. These findings support the existence of a positive feedback between surface albedo, vegetation, and precipitation starting already on the mesoscale.
We are convinced that quantitative analyses of changes in the water cycle have to be performed with the next generation of mesoscale models, with their sophisticated representations of vegetation on the convection-permitting scale. These simulations provide improved analyses of the impact of the plantations for three reasons: 1) more accurate simulation of the modification of the surface fluxes in the EBC, 2) resolution of the changes in flow patterns and the development of convergence zones as a result of land surface heterogeneity, and 3) the elimination of severe errors due to the parameterization of deep convection.
These effects scale with the size of the plantation. At a scale of 10 000 km2, precipitation enhancement is evident both in Oman and the Sonora. In the future, the impact on local climate dependent on plantation size, orientation, and shape will be studied in more detail with respect to convection initiation. The sensitivity of the model results could be investigated by ensemble runs using different model physics and resolutions. Detailed understanding and correct modeling of the transpiration of the plants in combination with precise irrigation management could even provide us with the ability to partially control for local climate effects. This type of afforestation could then be considered a form of climate engineering, which we define as climate engineering by afforestation.
Acknowledgments
This work was supported by the Stiftung Energieforschung and Jatropha Solutions, Baden-Württemberg, Germany. A part of the study was performed during a research stay at the Cooperative Institute for Research in Environmental Sciences of the University of Colorado Boulder. Further support was provided by the German Research Foundation (DFG) Research Unit (FOR) 1695. Figures 15–17 were produced with the NCAR VAPOR software (see http://www.vapor.ucar.edu). We appreciate the access to ECMWF analysis data.
REFERENCES
Avissar, R., and Y. Liu, 1996: A three-dimensional numerical study of shallow convective clouds and precipitation induced by land-surface forcing. J. Geophys. Res., 101, 7499–7518, doi:10.1029/95JD03031.
Bauer, H.-S., T. Weusthoff, M. Dorninger, V. Wulfmeyer, T. Schwitalla, T. Gorgas, M. Arpagaus, and K. Warrach-Sagi, 2011: Predictive skill of a subset of the D-PHASE multi-model ensemble in the COPS region. Quart. J. Roy. Meteor. Soc., 137, 287–305, doi:10.1002/qj.715.
Becker, K., V. Wulfmeyer, T. Berger, J. Gebel, and W. Münch, 2013: Carbon farming in hot, dry coastal areas: An option for climate change mitigation. Earth Syst. Dyn., 3, 1221–1258, doi:10.5194/esdd-3-1221-2012.
Bougeault, P., J. Noilhan, P. Lacarrère, and P. Mascart, 1991: An experiment with an advanced surface parameterization in a mesobeta-scale model. Part I: Implementation. Mon. Wea. Rev., 119, 2358–2373, doi:10.1175/1520-0493(1991)119<2358:AEWAAS>2.0.CO;2.
Brovkin, V., M. Claussen, V. Petoukhov, and A. Ganopolski, 1998: On the stability of the atmosphere–vegetation system in the Sahara/Sahel region. J. Geophys. Res., 103, 31 613–31 624, doi:10.1029/1998JD200006.
Charney, J. G., 1975: Dynamics of deserts and drought in the Sahel. Quart. J. Roy. Meteor. Soc., 101, 193–202, doi:10.1002/qj.49710142802.
Chen, F., and J. Dudhia, 2001: Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity. Mon. Wea. Rev., 129, 569–585, doi:10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2.
Chen, F., and Coauthors, 1996: Modeling of land surface evaporation by four schemes and comparison with FIFE observations. J. Geophys. Res.,101D, 7251–7268, doi:10.1029/95JD02165.
Claussen, M., 1994: On coupling global biome models with climate models. Climate Res., 4, 203–221, doi:10.3354/cr004203.
Cutrim, E., D. W. Martin, and R. Rabin, 1995: Enhancement of cumulus clouds over deforested lands in Amazonia. Bull. Amer. Meteor. Soc., 76, 1801–1805, doi:10.1175/1520-0477(1995)076<1801:EOCCOD>2.0.CO;2.
Dalu, G. A., R. A. Pielke, M. Baldi, and X. Zeng, 1996: Heat and momentum fluxes induced by thermal inhomogeneities. J. Atmos. Sci., 53, 3286–3302, doi:10.1175/1520-0469(1996)053<3286:HAMFIB>2.0.CO;2.
Dudhia, J., 1989: Numerical study of convection observed during the Winter Monsoon Experiment using a mesoscale two-dimensional model. J. Atmos. Sci., 46, 3077–3107, doi:10.1175/1520-0469(1989)046<3077:NSOCOD>2.0.CO;2.
Edenhofer, O., and Coeditors, 2012: Meeting report of the Intergovernmental Panel on Climate Change Expert Meeting on Geoengineering. Potsdam Institute for Climate, 99 pp.
Fairless, D., 2007: Biofuel: The little shrub that could—maybe. Nature, 449, 652–655, doi:10.1038/449652a.
Feldmann, H., B. Früh, G. Schädler, H.-J. Panitz, K. Keuler, D. Jacob, and P. Lorenz, 2008: Evaluation of the precipitation for southwestern Germany from high resolution simulations with regional climate models. Meteor. Z., 17, 455–465, doi:10.1127/0941-2948/2008/0295.
Friedl, M. A., D. K. McIver, J. C. F. Hodges, X. Y. Zhang, D. Muchoney, A. H. Strahler, C. E. Woodcock, and C. Schaaf, 2002: Global land cover mapping from MODIS: Algorithms and early results. Remote Sens. Environ., 83, 287–302, doi:10.1016/S0034-4257(02)00078-0.
Glenn, E. P., C. N. Hodges, H. Lieth, R. A. Pielke, and L. Pitelka, 1992: Growing halophytes to remove carbon from the atmosphere. Environment, 34, 40–43, doi:10.1080/00139157.1992.9931438.
Greve, P., K. Warrach-Sagi, and V. Wulfmeyer, 2013: Evaluating soil water content in a WRF-Noah downscaling experiment. J. Appl. Meteor. Climatol., 52, 2312–2327, doi:10.1175/JAMC-D-12-0239.1.
Gupta, R. K., T. S. Prasad, and D. Vijayan, 2002: Estimation of roughness length and sensible heat flux from WiFS and NOAA AVHRR data. Adv. Space Res.,29, 33–38, doi:10.1016/S0273-1177(01)00624-X.
Hänel, H.-D., and F. J. Löpmeier, 1997: Bereitstellung von Parameterisierungsverfahren zur Bestimmung der regionalen Verdunstung verschiedener Landoberflächen (Approaches for the parameterization of regional evapotranspiration from different land surfaces). Deutscher Wetterdienst Braunschweig, 110 pp.
Haywood, J. M., A. Jones, N. Bellouin, and D. Stephenson, 2013: Asymmetric forcing from stratospheric aerosols impacts Sahelian rainfall. Nat. Climate Change, 3, 660–665, doi:10.1038/nclimate1857.
Henderson-Sellers, A., Z.-L. Yang, and R. E. Dickinson, 1993: The Project for Intercomparison of Land-surface Parameterization Schemes. Bull. Amer. Meteor. Soc., 74, 1335–1349, doi:10.1175/1520-0477(1993)074<1335:TPFIOL>2.0.CO;2.
Hohenegger, C., P. Brockhaus, C. S. Bretherton, and C. Schär, 2009: The soil moisture–precipitation feedback in simulations with explicit and parameterized convection. J. Climate, 22, 5003–5020, doi:10.1175/2009JCLI2604.1.
Holl, M. A., M. B. Gush, J. Hallowes, and D. Versfeld, 2007: Jatropha curcas in South Africa: An assessment of its water use and bio-physical potential. WRC Rep. 1497/1/07, Water Resources Commission, Pretoria, South Africa, 134 pp. [Available online at http://jatropha.pro/PDF%20bestanden/JatrophaWater%20South%20Africa.pdf.]
Hong, S.-Y., Y. Noh, and J. Dudhia, 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 2318–2341, doi:10.1175/MWR3199.1.
IUSS, 2007: World reference base for soil resources 2006: First update 2007. World Soil Resources Rep. 103, International Union of Soil Scientists Working Group WRB, FAO, Rome, Italy, 116 pp.
Kendon, E. J., N. M. Roberts, C. A. Senior, and M. J. Roberts, 2012: Realism of rainfall in a very high-resolution regional climate model. J. Climate, 25, 5791–5806, doi:10.1175/JCLI-D-11-00562.1.
Kleidon, A. K., K. Fraedrich, and M. Heimann, 2000: A green planet versus a desert world: Estimating the maximum effect of vegetation on the land surface climate. Climatic Change, 44, 471–493, doi:10.1023/A:1005559518889.
Lenton, T. M., and N. E. Vaughan, 2009: The radiative forcing potential of different climate geoengineering options. Atmos. Chem. Phys., 9, 5539–5561, doi:10.5194/acp-9-5539-2009.
Li, Z., B.-L. Lin, X. Zhao, M. Sagisaka, and R. Shibazaki, 2010: System approach for evaluating the potential yield and plantation of Jatropha curcas L. on a global scale. Environ. Sci. Technol., 44, 2204–2209, doi:10.1021/es903004f.
Lyons, T. J., H. Xinmei, P. Schwerdtfeger, J. M. Hacker, I. J. Foster, and R. C. G. Smith, 1993: Land–atmosphere interaction in a semiarid region: The bunny fence experiment. Bull. Amer. Meteor. Soc., 74, 1327–1334, doi:10.1175/1520-0477(1993)074<1327:LIIASR>2.0.CO;2.
Mahfouf, J. F., E. Richard, and P. Mascart, 1987: The influence of soil and vegetation on the development of mesoscale circulations. J. Climate Appl. Meteor., 26, 1483–1495, doi:10.1175/1520-0450(1987)026<1483:TIOSAV>2.0.CO;2.
Mahmood, R., and Coauthors, 2014: Land cover changes and their biogeophysical effects on climate. Int. J. Climatol., 34, 929–953, doi:10.1002/joc.3736.
Mahrt, L., and M. Ek, 1984: The influence of atmospheric stability on potential evaporation. J. Climate Appl. Meteor.,23, 222–234, doi:10.1175/1520-0450(1984)023<0222:TIOASO>2.0.CO;2.
McAlpine, C. A., and Coauthors, 2010: More than CO2: A broader picture for managing climate change and variability to avoid ecosystem collapse. Curr. Opin. Environ. Sustainability, 2, 334–336, doi:10.1016/j.cosust.2010.10.001.
Metz, B., O. R. Davidson, P. R. Bosch, R. Dave, and L. A. Meyer, Eds., 2007: Climate Change 2007: Mitigation of Climate Change. Cambridge University Press, 862 pp.
Milovac, J., F. Späth, K. Warrach-Sagi, A. Behrendt, and V. Wulfmeyer, 2013: Evaluation of WRF boundary-layer parameterization schemes and land-surface models by comparing an ensemble with 44 members with scanning water vapor DIAL measurements. TR32-HOBE Int. Symp., Bonn, Germany, Transregional Collabrative Research Center 32–Center for Hydrology. [Available online at http://tr32meeting.uni-koeln.de/images/abstract/s3/TR32-HOBE-S3-Wednesday-1140-1200_Milovac.pdf.]
Mlawer, E. J., S. J. Taubman, P. D. Brown, M. J. Iacono, and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102D, 16 663–16 682, doi:10.1029/97JD00237.
Morrison, H., G. Thompson, and G. Tatarskii, 2009: Impact of cloud microphysics on the development of trailing stratiform precipitation in a simulated squall line: Comparison of one- and two-moment schemes. Mon. Wea. Rev., 137, 991–1007, doi:10.1175/2008MWR2556.1.
Perlin, N., and P. Alpert, 2001: Effects of land use modification on potential increase of convection: A numerical mesoscale study over south Israel. J. Geophys. Res., 106 (D19), 22 621–22 634, doi:10.1029/2000JD900804.
Pielke, R. A., G. A. Dalu, J. A. Snook, T. J. Lee, and T. G. F. Kittel, 1991: Nonlinear influence of mesoscale land use on weather and climate. J. Climate, 4, 1053–1069, doi:10.1175/1520-0442(1991)004<1053:NIOMLU>2.0.CO;2.
Pielke, R. A., J. Adegoke, A. Beltran-Przekurat, C. A. Hiemstra, J. Lin, U. S. Nair, D. Niyogi, and T. E. Nobis, 2007: An overview of regional land use and land cover impacts on rainfall. Tellus, 59B, 587–601, doi:10.1111/j.1600-0889.2007.00251.x.
Polcher, J., and K. Laval, 1994: A statistical study of the regional impact of deforestation on climate in the LMD GCM. Climate Dyn., 10, 205–229, doi:10.1007/BF00208988.
Prentice, I. C., W. Cramer, S. P. Harrison, R. Leemans, R. A. Monserud, and A. M. Solomon, 1992: A global biome model based on plant physiology and dominance, soil properties and climate. J. Biogeogr., 19, 117–134, doi:10.2307/2845499.
Rabin, R. M., and D. W. Martin, 1996: Satellite observations of shallow cumulus coverage over the central United States: An exploration of land use impact on cloud cover. J. Geophys. Res., 101, 7149–7155, doi:10.1029/95JD02891.
Righelato, R., and D. Spracklen, 2007: Carbon mitigation by biofuels or by saving and restoring forests? Science, 317, 902, doi:10.1126/science.1141361.
Rotach, M. W., and Coauthors, 2009a: MAP D-PHASE: Real-time demonstration of weather forecast quality in the Alpine region. Bull. Amer. Meteor. Soc., 90, 1321–1336, doi:10.1175/2009BAMS2776.1.
Rotach, M. W., and Coauthors, 2009b: Supplement to MAP D-PHASE: Real-time demonstration of weather forecast quality in the Alpine region: Additional applications of the D-Phase datasets. Bull. Amer. Meteor. Soc.,90, S28–S32, doi:10.1175/2009BAMS2776.2.
Schwitalla, T., G. Zängl, H.-S. Bauer, and V. Wulfmeyer, 2008: Systematic errors of QPF in low-mountain regions. Meteor. Z., 17, 903–919, doi:10.1127/0941-2948/2008/0338.
Schwitalla, T., H.-S. Bauer, V. Wulfmeyer, and F. Aoshima, 2011: High-resolution simulation over central Europe: Assimilation experiments with WRF 3DVAR during COPS IOP9c. Quart. J. Roy. Meteor. Soc., 137, 156–175, doi:10.1002/qj.721.
Spracklen, D. V., S. R. Arnold, and C. M. Taylor, 2012: Observations of increased tropical rainfall preceded by air passage over forests. Nature, 489, 282–285, doi:10.1038/nature11390.
Taylor, G. I., 1916: The skin friction of the wind on the earth's surface. Proc. Roy. Soc. London, 92A, 196–199.
Vautard, R., and Coauthors, 2013: The simulation of European heat waves from an ensemble of regional climate models within the EURO-CORDEX project. Climate Dyn., 41, 2555–2575, doi:10.1007/s00382-013-1714-z.
Viña, A. H., M. Geoffrey, and A. A. Gitelson, 2004: Satellite monitoring of vegetation dynamics: Sensitivity enhancement by the wide dynamic range vegetation index. Geophys. Res. Lett.,31, L04503, doi:10.1029/2003GL019034.
Warner, T. T., 2004: Desert Meteorology. Cambridge University Press, 595 pp.
Warrach-Sagi, K., T. Schwitalla, V. Wulfmeyer, and H.-S. Bauer, 2013: Evaluation of a climate simulation based on the WRF–NOAH model system: Precipitation in Germany. Climate Dyn., 41, 755-774, doi:10.1007/s00382-013-1727-7.
Weaver, C. P., and R. Avissar, 2001: Atmospheric disturbances caused by human modification of the landscape. Bull. Amer. Meteor. Soc., 82, 269–281, doi:10.1175/1520-0477(2001)082<0269:ADCBHM>2.3.CO;2.
Weisman, M. L., C. Davis, W. Wang, K. W. Manning, and J. B. Klemp, 2008: Experiences with 0–36-h explicit convective forecasts with the WRF-ARW model. Wea. Forecasting, 23, 407–437, doi:10.1175/2007WAF2007005.1.
Weusthoff, T., F. Ament, M. Arpagaus, and M. W. Rotach, 2010: Assessing the benefits of convection permitting models by fuzzy verification: Examples from MAP D-PHASE. Mon. Wea. Rev., 138, 3418–3433, doi:10.1175/2010MWR3380.1.
Wulfmeyer, V., and Coauthors, 2008: The Convective and Orographically Induced Precipitation Study: A research and development project of the World Weather Research Program for improving quantitative precipitation forecasting in low-mountain regions. Bull. Amer. Meteor. Soc., 89, 1477–1486, doi:10.1175/2008BAMS2367.1.
Wulfmeyer, V., and Coauthors, 2011: The Convective and Orographically Induced Precipitation Study (COPS): The scientific strategy, the field phase, and first highlights. Quart. J. Roy. Meteor. Soc.,137, 3–30, doi:10.1002/qj.752.
