Hanna Meyer

Environmental Changes Affecting the Andes of Ecuador

Global terrestrial biodiversity is strongly affected by expanding land use, climate change and nitrogen deposition. This holds especially true for tropical forests which already show large changes due mainly to land use activities. The extent of land use in Ecuador has increased considerably during the last century. An extensive network of primary and secondary roads now opens up most of the western and central areas of the country, while parts of the Oriente have been converted into protected areas. Concerning climate change warming is predicted to be moderate for western Ecuador, while the eastern part of the country will suffer from rising temperatures that will affect a floristic region harbouring one of the global diversity hotspots for vascular plant species. Changes in precipitation are expected to be spatially much less cohesive, with increasing and decreasing amounts of precipitation being unevenly distributed throughout the Andes. The spatial distribution and temporal dynamics of precipitation and wind also regulate the deposition of rainwater-dissolved matter in the mountain ecosystem which results from biomass burning in Amazonia. In this chapter, our current knowledge as to the past development of these major threats of the ecosystem will be discussed focusing on the study area South Ecuador.

Mapping Daily Air Temperature for Antarctica Based on MODIS LST

Spatial predictions of near-surface air temperature ( T a i r ) in Antarctica are required as baseline information for a variety of research disciplines. Since the network of weather stations in Antarctica is sparse, remote sensing methods have large potential due to their capabilities and accessibility. Based on the MODIS land surface temperature (LST) data, T a i r at the exact time of satellite overpass was modelled at a spatial resolution of 1 km using data from 32 weather stations. The performance of a simple linear regression model to predict T a i r from LST was compared to the performance of three machine learning algorithms: Random Forest (RF), generalized boosted regression models (GBM) and Cubist. In addition to LST, auxiliary predictor variables were tested in these models. Their relevance was evaluated by a Cubist-based forward feature selection in conjunction with leave-one-station-out cross-validation to reduce the impact of spatial overfitting. GBM performed best to predict T a i r using LST and the month of the year as predictor variables. Using the trained model, T a i r could be estimated with a leave-one-station-out cross-validated R 2 of 0.71 and a RMSE of 10.51 ∘ C. However, the machine learning approaches only slightly outperformed the simple linear estimation of T a i r from LST ( R 2 of 0.64, RMSE of 11.02 ∘ C). Using the trained model allowed creating time series of T a i r over Antarctica for 2013. Extending the training data by including more years will allow developing time series of T a i r from 2000 on.

Projecting land-use and land-cover changes in a tropical mountain forest of Southern Ecuador

Land-use and land-cover changes (LULCC) affect local climate. Human-induced deforestation is a common phenomenon of LULCC. This also holds true for the biodiversity hotspot in the Andes of Ecuador. This study assesses the possibility to project LULCC to future time steps with a focus on deforestation in the San Francisco valley in South Ecuador. A business-as-usual scenario based on two Landsat scenes from 1987 and 2001 was created to project LULCC until 2006. The uncertainty assessment indicated the difficulty of projecting human impact on the ecosystem since circumstances of LULCC in the study area cannot be assumed to stay invariant. The projection performs better than a naive model relying on slope as its suitability factor.

Assessing pasture quality and degradation status using hyperspectral imaging: a case study from western Tibet

Alpine grasslands on the Tibetan Plateau (TP) are suffering from pasture degradation induced by over-grazing, climate change and improper livestock management. Meanwhile, the status of pastures is largely unknown especially in poor accessible western parts on the TP. The aim of this case study was to assess the suitability of hyperspectral imaging to predict quality and amount of forage on the western TP. Therefore, 18 ground- based hyperspectral images taken along two transects on a winter pasture were used to estimate leaf chlorophyll content, photosynthetic-active vegetation cover (PV) and proportion of grasses. For calibration and validation purposes, chlorophyll content of 20 grass plants was measured in situ. From the images reference spectra of grass and non-grass species were collected. PV was assessed from similarity of images to mean vegetation spectra using spectral angle mapper and threshold classifications. A set of 48 previously published hyperspectral vegetation indices (VI) was used as predictors to estimate chlorophyll content and to discriminate grass and non-grass pixels. Separation into grass and non-grass species was performed using partial least squares (PLS) discriminant analysis and chlorophyll content was estimated with PLS regression. The accuracy of the models was assessed with leave-one-out cross validation and normalised root mean square errors (nRMSE) for chlorophyll and contingency matrices for grass classification and total PV separation. Highest error rates were observed for discrimination between vegetated and non-vegetated parts (Overall accuracy = 0.85), whilst accuracies of grass and non grass separation (Overall accuracy = 0.98) and chlorophyll estimation were higher (nRMSE = 10.7).

Revealing the potential of spectral and textural predictor variables in a neural network-based rainfall retrieval technique

ABSTRACT Estimating rainfall areas and rates from geostationary satellite images has the opportunity of both, a high spatial and a high temporal resolution which cannot be achieved by other satellite-based systems until now. Most recent retrieval techniques are solely based on spectral channels of the satellites. These retrievals can be classified as ‘purely pixel-based’ because no information about the neighbourhood pixels is included. Assuming that precipitation is highly correlated with cloud processes and therefore with cloud texture, textural information derived from the neighbourhood of a pixel might give valuable information about the cloud type and hence about a respective probability of the rainfall rate. To study the potential of textural variables to improve optical rainfall retrieval techniques, rainfall areas and rainfall rates were estimated over Germany for the year 2010 using a neural network approach. In addition to the spectral predictor variables from Meteosat Second Generation (MSG), different Grey Level Co-occurance Matrix (GLCM) based textural variables were calculated from all MSG channels. Using recursive feature selection, models were trained and their performance was compared to spectral-only models. Contrary to the expectations, the performance of the models did not increase when textural information was included.

Current Provisioning Services : Pasture Development and Use, Weeds (Bracken) and Management

This chapter reports on the historical expansion and current state of the pastures in the Rio San Francisco valley. Its major part is inhabited by the Mestizos, who do not have a long-standing pasture tradition. Three types of pastures were identified by the dominant grass species: the “pastos azules” (Holcus lanatus), the Yaragua pastures (Melinis minutiflora) and the dominating “pastos mieles” (Setaria sphacelata). The peculiarities, species composition, soil dynamics and agricultural values of these pastures are discussed. Except for the pastos azules on small flattenings in the otherwise steep slopes of the valley, pastures in the area suffer from invasion by aggressive weeds, mainly the tropical bracken fern. Abandonment of pastures is fostered by the use of fire to combat weeds and stimulate grass growth. This type of low-yield pasture farming is not sustainable. The earnings of livestock farming are not sufficient for subsistence. Diversification of the income portfolio is necessary.

Impacts of Local Land-Use Change on Climate and Hydrology

Land-use change has a potentially large impact on local water resources and climatic conditions in montane rainforest ecosystems of the Andes. Based on local meteorological observations and site-specific simulation studies involving a coupled hydrological model and a soil–vegetation–atmosphere transfer scheme, we are able to predict likely changes of water and energy fluxes for different land-use categories. To anticipate the effect of future land-use change on the water and energy budgets of the study area, we use results of statistically derived land-use scenarios and a coupled plot scale model representing the dominant land-use types for further upscaling. After assessing the impact of land-use change on ecosystem services we conclude that climate regulation will be decreasing due to a likely increase in drought vulnerability and that the discharge will remain stable or even slightly increase, thereby positively effecting provisioning and regulating hydrological services.

Measuring pasture degradation on the Qinghai-Tibet Plateau using hyperspectral dissimilarities and indices

Despite that relevance of pasture degradation on the Qinghai-Tibet Plateau (QTP) is widely postulated, its extent is still unknown. However, livestock grazing is widely accepted as a major factor. This study investigated spectral differences of vegetation patterns along gradients of grazing intensities using plot-based hyperspectral measurements. The measurements were used to define spectral indicators for pasture degradation, which were applied to map asserted proxies for degradation from satellite images. For this purpose, hyperspectral measurements were taken at 11 sites on the north-eastern QTP using a transect design from heavy grazing and therefore asserted degradation near the settlement to less degradation with increasing distance. Potential spectral indicators for degradation were derived from the spectra by calculating the size of continuum removed absorption features and narrow-band indices (NBI). They were compared between degraded and less degraded plots. Linear regressions between proxies and each of the potential spectral indicators were calculated to assess its predictive power. The findings were transferred to larger scales by applying the indicators on two WorldView-2 (WV-2) scenes. Spectral differences between degraded and less degraded plots were obvious regarding a wide range of tested indicators. Several NBIs were considered as good indicators for vegetation cover and species numbers. WV-2 images could be successfully classified into vegetation cover whilst the estimation of species numbers was afflicted with uncertainties. The results demonstrate the potential to estimate degradation proxies using spectrometer measurements and satellite data. Applying these techniques will contribute to a better estimation of spatial degradation patterns on the QTP.