Hi Sri
2) If there is a DAC, then the signal preceding it must be in digital form. So, when you refer to "analog signal is present at the output of Modulation block" then this "analog signal" is not truly in analog form, but is a digital representation of the analog signal. Otherwise it does not make sense.
I hope that you understand that when you create a signal in a digital signal processing, the signal is only a digital representation using numbers. These numbers must be converted to an analog voltage, current or power at some point in order to be transmitted and this is the job of the DAC and circuits following the DAC.
1) signal processing for a transmitter is done at baseband rather than at the transmitter output frequency because the DSP (digital signal processor) electronics would be too expensive and too power hungry (and sometimes not feasible) if the DSP is done at the transmitter output frequency. To make a transmitter in a practical and economical way, we create the signal at baseband knowing that there are various ways to then move this baseband signal up to the transmitter operating frequency. One way to move up to the transmitter output frequency is called "translation". The result of translation is that the baseband spectrum is moved up in frequency without changing the width or relative amplitude of the spectrum. Translation is done using a mixer and a local oscillator and is usually called "up conversion" or "up mixing" or "direct modulation" and some other names. It is most common now that signals are translated using IQ up conversion, wherein the I (in phase) information follows one upconversion path while the Q (quadrature) information follows a different upconversion path. This is also a method of translation but it has the advantage of allowing all of the amplitude and phase information to be controlled at the cost of more complexity in the up conversion circuitry. IQ up conversion is found in most cell phones, WiFi, and other digital wireless systems.