Impact on simulated surface energy fluxes and temperature

Abstract. Land surface models are essential parts of climate and weather models. The
widely used Noah-MP land surface model requires information on the leaf area
index (LAI) and green vegetation fraction (GVF) as key inputs of its
evapotranspiration scheme. The model aggregates all agricultural areas into
a land use class termed “cropland and pasture”. In a previous study we
showed that, on a regional scale, the GVF has a bimodal distribution formed by
two crop groups differing in phenology and growth dynamics: early-covering
crops (ECC; e.g., winter wheat, winter rapeseed, winter barley) and late-covering crops (LCC; e.g., corn, silage maize, sugar beet). That result can
be generalized for central Europe. The present study quantifies the effect
of splitting the land use class cropland and pasture of Noah-MP into ECC and LCC on surface energy fluxes and temperature. We further studied the
influence of increasing the LCC share, which in the study area (the
Kraichgau region, southwest Germany) is mainly the result of heavily
subsidized biomass production, on energy partitioning at the land surface.
We used the GVF dynamics derived from high-resolution (5 m × 5 m) RapidEye
satellite data and measured LAI data for the simulations. Our results
confirm that the GVF and LAI strongly influence the partitioning of surface
energy fluxes, resulting in pronounced differences between simulations of ECC and LCC. Splitting up the generic crop into ECC and LCC had the
strongest effect on land surface exchange processes in July–August. During
this period, ECC are at the senescence growth stage or already harvested,
while LCC have a well-developed ground-covering canopy. The generic crop
resulted in humid bias, i.e., an increase in evapotranspiration by +0.5 mm d−1 (latent heat flux is 1.3 MJ m−2 d−1), decrease in sensible heat flux (H)
by 1.2 MJ m−2  d−1 and decrease in surface temperature by
−1 ∘C. The bias increased as the shares of ECC and LCC became
similar. The observed differences will impact the simulations of processes
in the planetary boundary layer. Increasing the LCC share from 28 % to 38 %
in the Kraichgau region led to a decrease in latent heat flux (LE) and a
heating up of the land surface in the early growing season. Over the second
part of the season, LE increased and the land surface cooled down by up to
1 ∘C.


The objectives of the present study were 1) to elucidate the extent to which surface energy fluxes    The GVF data required by the Noah-MP model were derived from high-resolution (5 m x 5 m) 152 RapidEye satellite data. As described by Imukova et al. (2015) the GVF data were calculated from 153 the Normalized Difference Vegetation Index (NDVI) computed from the red and near-infrared 154 bands of the satellite images. The relationship between GVF and NDVI was established by linear 155 regression using ground truth measurements. GVF maps were derived in a monthly resolution.      190 We firstly quantified the extent to which ECC and LCC differ with regard to their energy and water 191 fluxes, surface (TS) and soil temperature (TG). For this, we performed one simulation for each 192 crop group using the mean LAI and the mean GVF dynamics observed during the two growing 193 seasons (see Table 2 and Table 3). that of ECC and LCC, we compared the following two simulation runs:

197
Run 1: Noah-MP was forced with the GVF and LAI dynamics of the generic crop (Table 2 and 198 Table 3). Accordingly, in this simulation, we first computed the weighted mean of the vegetation   (Table 4). Compared with LCC, the higher latent 225 heat fluxes of ECC in May and June resulted in a cooler land surface, on average by -2.6°C and -226 1.0°C, respectively (Table 4). From July to August the situation was reversed: because latent heat 227 fluxes of ECC are distinctly lower than that of LCC, the surface temperature at ECC sites was up 228 to 4°C warmer than at LCC sites ( Figure 2).

230
The mean difference in daily ground heat flux between ECC and LCC during the growing season 231 ranged between -0.2 MJ m -2 and 0.2 MJ m -2 (Table 4). Also for the ground heat flux, the smallest 232 difference between both crops types was observed in June (0.05 MJ m -2 ).

292
Increasing the LCC fraction from 28% to 38% mainly led to changes in LE and H fluxes (Table   (Table 6). The potential increase of the LCC fraction (driven by the high demand for biogas and forage 327 production) leads to significant changes in the partitioning of the energy fluxes at the croplands.

328
In recent years the total area under maize in Germany has more than doubled. This corresponds to 329 an approximately 10% increase of the LCC fraction for the study region. In the early vegetation

21
The results presented above apply to the ECC-LCC ratio within our study area. What can we expect 346 in agricultural landscapes with different ECC-LCC ratios? The ECC-LCC ratio has nearly no effect 347 on energy partitioning in June, whereas in May, July and August its influence on the turbulent 348 fluxes is pronounced (Figure 6). The weak effect in June is because, during this period, the LAI 349 and GVF of ECC and LCC are similar (Figure 8). In the other months, however, the ECC-LCC  (Table 7). Moreover, different ECC-LCC ratios will also affect 355 the above-mentioned humid bias of the generic crop parameterization (Figure 7).     partitioning at the land surface on the ABL evolution on a diurnal scale. They observed that LE 417 simulated by Noah-MP was more than 50% lower than that simulated by Noah. As expected, a 418 lower LE resulted in a drier ABL. The ABL evolution and its features strongly influence the 419 initiation of convection and cloud formation as well as the location and strength of precipitation.

420
For instance, drier and higher ABL would yield a higher lifting condensation level, leading to 421 higher clouds and a higher probability of convective precipitation.